<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by ARTURO on Medium]]></title>
        <description><![CDATA[Stories by ARTURO on Medium]]></description>
        <link>https://medium.com/@arturo-ai?source=rss-d736cf5bbc00------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*wrXyqRoCzyJT9TpwCsVv-g.jpeg</url>
            <title>Stories by ARTURO on Medium</title>
            <link>https://medium.com/@arturo-ai?source=rss-d736cf5bbc00------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 14:54:05 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@arturo-ai/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[The Arturo Difference: The Most Advanced Technology Across P&C Insurance — Arturo]]></title>
            <link>https://arturo-ai.medium.com/the-arturo-difference-the-most-advanced-technology-across-p-c-insurance-arturo-f83eef273eab?source=rss-d736cf5bbc00------2</link>
            <guid isPermaLink="false">https://medium.com/p/f83eef273eab</guid>
            <category><![CDATA[technology]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[insurance]]></category>
            <category><![CDATA[imagery]]></category>
            <dc:creator><![CDATA[ARTURO]]></dc:creator>
            <pubDate>Fri, 19 Feb 2021 16:35:50 GMT</pubDate>
            <atom:updated>2021-03-08T20:18:13.959Z</atom:updated>
            <content:encoded><![CDATA[<h3>The Arturo Difference: The Most Advanced Technology Across P&amp;C Insurance — Arturo</h3><p>At Arturo, we strive to offer P&amp;C Insurance Carriers the most progressive technology that advances the insurance industry forward. We do this by constantly seeking additional imagery sources, strategic partnerships and building cutting-edge technology to enhance our process and push the boundaries on what we offer clients.</p><p>What makes our technology so progressive? Continue reading this article to learn about Arturo’s key technology differentiators across our data, our process, our delivery, and our future.</p><h3>Multi-source Imagery</h3><p>To improve decision making across quotes, claims, and underwriting, Insurance Carriers need the most up-to-date property data. To attain the most accurate property data analytics, utilizing multi-source imagery across all categories of sources is the best option. That is exactly why we have worked so diligently to support multiple image providers. Arturo utilizes multiple sources across aerial, satellite, stratospheric balloons, drones, and ground level. This allows us to deliver the most up-to-date, complete, and comprehensive property coverage across the insurance industry. You can learn more about our strategic partners .</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*vuL6RShTKNDevYpb.jpg" /></figure><h3>Historical Claims and Inspections</h3><p>Originating from the insurance industry, Arturo’s unparalleled access to claims and inspection history, across a variety of Insurance Carriers, sets us apart. Arturo offers higher-level predictive models that can only be generated from carrier-specific internal data, giving insurers critical property analytics that can’t be found anywhere else.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*rPnP76ifSOdf1Fqt.jpg" /></figure><h3>Temporal Resolution</h3><p>Temporal Resolution is a geospatial term that expresses how often a particular area’s imagery is re-captured and reprocessed. The higher the refresh rate, the more up-to-date the imagery. Arturo is the only provider to deliver temporal property insights when Insurance Carriers need them most, even in areas where getting imagery can vary by season. You can learn more about the importance of temporal resolution by reading this <a href="https://www.arturo.ai/white-paper/how-to-evaluate-imagery-providers/">whitepaper </a>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*bevcYMlF0bkEP-fj.jpg" /></figure><h3>Confidence Scores</h3><p>Oftentimes the data that’s delivered to Insurance Carriers from data aggregators is inaccurate or incomplete. Arturo provides Confidence Scores, per property attribute, so you know you can trust the data. We pride ourselves in making an effort to incorporate critical ground truth data from Insurance Carriers to significantly improve the accuracy, precision, and recall of our models.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*NIvG63ekfho1Cq5L.jpg" /></figure><h3>On-demand Property Analytics</h3><p>There’s no preprocessing here at Arturo. Our technology enables Insurance Carriers to meet internal response SLAs “on-the-fly” without any imagery pre-fetching or preprocessing. Arturo’s on-demand property analytics dramatically reduces costs and improves price while removing the temporal problems of false-positive returns that undermine sound Underwriting decisions.</p><p>By harnessing a combination of efficient processing pipelines, Arturo is the only provider to offer on-demand analytics for more than 50 highly accurate property characteristics and predictive indicators.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*WDiql73nCzIp2Z_L.jpg" /></figure><h3>Feedback Loops</h3><p>Getting feedback from Insurance Carriers is a top priority at Arturo. We ensure that our clients are provided a feedback loop per API interaction via our product deployment. This continuous feedback powers the enhancement of our data models.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*xkwDcD52wvhkeOyy.jpg" /></figure><h3>Portfolio Book Processing</h3><p>Arturo’s portfolio processing capabilities are unrivaled with full geographic reach whether your analytical requirements are in-country, regional, or neighborhood-based. Our proven ability to process millions of addresses within days provides significant benefits to Insurance Carriers managing portfolio risks, major peril events, and rating scenario portfolio risks.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*TfhfoXwdSl24zirD.jpg" /></figure><h3>Persona-based APIs</h3><p>No need to change your internal best practices to get the property characteristics you need. Arturo delivers highly configurable APIs so Insurance Carriers get only the characteristics you want in under five seconds (for the average residential property). Our persona-based APIs provide the data you need in industry-standard, multiple formats that are easily ingested into your current business workflows and processes. With Arturo, you can be up and running in days, not weeks.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*61rD3Fz1zdysJbki.jpg" /></figure><h3>Driving Innovation</h3><p>Client-centric innovation is at the heart of everything we do at Arturo. We push ourselves to offer insurers on-demand insights without preprocessing and we developed an unprecedented ability to process millions of property records in 48 hours. Arturo’s willingness and agility to integrate feedback, include additional characteristics and improve our outputs, sets us apart.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*fSOxiHkYMAnduWDU.jpg" /></figure><p>If you’re interested in learning more or want to see our technology in action, click <a href="https://www.arturo.ai/technology/">here</a>.</p><p><em>Originally published at </em><a href="https://www.arturo.ai/the-arturo-difference-the-most-advanced-technology-across-pc-insurance/"><em>https://www.arturo.ai</em></a><em> on February 19, 2021.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f83eef273eab" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Sky’s the Limit: Why Arturo Continues to Partner with New and Novel Imagery Providers — Arturo]]></title>
            <link>https://arturo-ai.medium.com/the-skys-the-limit-why-arturo-continues-to-partner-with-new-and-novel-imagery-providers-arturo-e5ac26f870e1?source=rss-d736cf5bbc00------2</link>
            <guid isPermaLink="false">https://medium.com/p/e5ac26f870e1</guid>
            <category><![CDATA[imagery]]></category>
            <dc:creator><![CDATA[ARTURO]]></dc:creator>
            <pubDate>Wed, 16 Dec 2020 15:42:06 GMT</pubDate>
            <atom:updated>2021-03-08T20:16:20.149Z</atom:updated>
            <content:encoded><![CDATA[<h3>The Sky’s the Limit: Why Arturo Continues to Partner with New and Novel Imagery Providers — Arturo</h3><p>At Arturo, our team is constantly seeking partnerships with the most cutting edge providers across technology, data, and imagery to deliver the best property analytics to Insurance Carriers. We utilize multi-source imagery to enhance our process and push the boundaries on what we offer clients. We strive towards embracing as much data as possible because we know that one data provider can not solve all of the challenges Insurance Carriers face today.</p><p>Arturo has partnered with leaders across the imagery landscape including aerial, satellite, drone and ground level. Recently we have partnered with two up and coming players in the imagery space, Near Space Labs and Urban Sky, to derive imagery analytics from Stratospheric Balloons.</p><p>In this article we dive into what stratospheric balloon imagery is, why we are now partnering with balloon imagery providers, the interesting opportunity they present for Insurance Carriers, and hear directly from the technology leaders Near Space Labs and Urban Sky.</p><h3>What are Stratospheric Balloons?</h3><p>Stratospheric Balloons are manned or unmanned high-altitude balloons that are filled with helium and released into the stratosphere between 60,000–120,000 feet in elevation. Stratospheric balloons do not require an engine or fuel to fly. They fly with the air currents and their path is determined by very precise weather modeling. Balloons have been used for decades across fields of science, the military, and aerospace; recent advances in atmospheric modeling have opened the door to precise flight paths suitable for collecting overhead imagery of the earth.</p><p><strong>Using stratospheric balloons to capture imagery:</strong></p><p>When thinking about balloons in relation to other imagery sources, the altitude in which stratospheric balloons are flown is too low for satellites and too high for aircrafts. No pilots, no flight plans, no expensive jet fuel, no immense launch costs allow a much lower cost per pixel. So this kind of platform produces high-resolution imagery at a cost that is far less than satellites, aircrafts or drones.</p><p><strong>Frequency of imagery:</strong></p><p>Balloon sourced images can have a higher Temporal Resolution (or what is also known as the revisit-rate, expressing how often a particular area is re-captured and reprocessed). However, it is poignant to note that using stratospheric balloons as an imagery capture source is still in a very experimental phase. There is potential for balloons to fly very frequently, with the possibility of regular monthly coverage and daily to weekly coverage in certain areas of interest (for example, areas recently impacted by natural disasters). No other types of imagery providers collect images that often at such a detailed resolution.</p><p><strong>Resolution of imagery:</strong></p><p>Because the balloons are closer to earth, they are able to produce images with higher Spatial Resolution. This is a critical measurement for Insurance Carriers as it is the number one factor in determining a property data model’s accuracy. The Spatial Resolution is typically between 8cm — 30cm.</p><h3>How Stratospheric Balloons Imagery Enables Insurance Carriers</h3><p>Balloons offer an interesting opportunity for Insurance Carriers because they can be operated in a region of the atmosphere (15 to 45 km in altitude or roughly at 60,000ft), which is too low for satellites, too high for aircrafts. Or in other words they fall in an elevation that is between satellite and aerial imagery. Why does that matter?</p><p>Stratospheric Balloon Imagery produces near aerial-level Spatial Resolution with the ability to also deliver satellite level coverage (very large swaths in a short amount of time) and revisit (high frequency).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/0*NHVmobYTIYxXUIqZ.png" /></figure><p>Stratospheric Balloons have been around for a long time, and just recently they have been repurposed for imagery acquisition. Providers like Near Space Labs and Urban Sky are utilizing an existing aerial platform that has been understood and regulated for decades, for image collection by changing out the sensor package and camera system attached to the balloons. How brilliant!</p><p>Stratospheric balloons do not have the same FAA restrictions that airplanes have, they are far less costly, and they do not have any of the government resolution restrictions often imposed on satellite imagery. Additionally, due to the lower cost and ability to download large volumes of data, stratospheric balloon imagery will be more likely to collect imagery over ex-urban and rural areas.</p><p><strong>The benefits of stratospheric balloon imagery can be boiled down to these key points:</strong></p><ul><li>Higher revisit rate than satellite imagery with increased collection of frequently uncollected regions</li><li>Near aerial-level resolution</li><li>Increased geographical coverage opportunity</li><li>Less regulations than aerial or satellite imagery</li><li>Lower design, build, flight, and maintenance costs</li></ul><h3>Why Near Space Labs</h3><p>As a stratospheric imagery provider, Near Space Labs produces high resolution, high frequency imagery to its collection of overhead imagery for clients. Our <a href="https://arturo.ai/nearspace-labs/">partnership with Near Space Labs </a>increases the collection of aerial imagery for properties from two- to three-times per year to monthly/weekly; or even daily in the event of a natural disaster such as flooding, wildfires or coastal storm surge.</p><p>Near Space Labs, with their high-frequency observations and ability to include precise resolution, is a great addition to Arturo’s services. When a region or state is threatened by a natural disaster, we need to monitor properties in near real-time and through the partnership with Near Space Labs, Arturo is more responsive to our clients, whenever they need us.</p><p><strong>From Rema Matevosyan, CEO of Near Space Labs: </strong><em>“We are working with Arturo to solve a massive issue for P&amp;C insurance companies. Traditionally, they have lacked up-to-date, ground-truth imagery for underwriting, claims and assessing risk. Arturo’s platform adds a layer of intelligence to enable insurance customers to better understand how commercial and residential properties change over time and also for catastrophe-triaging.”</em></p><h3>Why Urban Sky</h3><p>Via their new stratospheric Microballoon ™, Urban Sky delivers broad area, high-resolution aerial imagery at a cost drastically less than satellites, aircraft or drones. The company’s novel invention, the high-altitude Microballoon, leverages advances in component miniaturization and first-of-its-kind material re-usability to exponentially reduce the cost of high-res remote sensing and weather-related data capture over broad areas, leading to more impactful data over more areas than ever before.</p><p><strong>From Andrew Antonio, CEO and Co-founder of Urban Sky:</strong> <em>“The stratosphere is virtually untapped, yet offers an enormous advantage for aerial imagery capture. It combines the best attributes of satellites and fixed wing aircraft, at a cost drastically less than either of those platforms. Arturo is one of the first application developers to see this, truly understand it, and invest in the space to the benefit of its own customers.”</em></p><h3>Looking Ahead: The Future of Stratospheric Balloon Imagery</h3><p>Looking to the future, the Arturo team sees the value of utilizing stratospheric balloon imagery for <a href="https://arturo.ai/cat-risk-modeling/">catososphertic risk modeling </a>and overall book monitoring.</p><p>Insurance Carriers want to be able to identify the risk of their overall book ‘before’ and ‘after’ a catastrophic event. With the FAA restrictions and regulations involved with aerial imagery, it can be difficult to acquire the necessary data to uplift CAT risk models. With the use of stratospheric balloons, Arturo will likely be able to access more images pre and post-event so Insurance Carriers get the property characteristics that assist with flood and wind related exposures.</p><p>For most Insurance Carriers, they want to be able to monitor their book of business by collecting property images four times a year (or once a quarter). That kind of data at the required spatial resolution, from a single provider, does not exist today. Stratospheric balloons can take flight once a month, giving new eyes on properties like never before.</p><p><em>“Arturo is changing the way the insurance sector leverages data by changing the way we look at the Earth, and we’re excited to partner with such an innovative team. Urban Sky is commercializing an entirely new type of data, and Arturo was first through the door to evaluate our products and imagine a future made better by their existence,” </em>states Andrew Antonio, CEO of Urban Sky.</p><p>The Arturo team is always pushing to find the best data available to empower our clients. The ability to ingest from myriad data sources is at the foundation of how we provide the best and most timely property characteristics possible. We are thrilled to bring on partners like Urban Sky and Near Space labs to push the boundaries on what we can offer Insurance Carriers. Learn more about our partners .</p><p><em>Originally published at </em><a href="https://www.arturo.ai/the-skys-the-limit-why-arturo-continues-to-partner-with-new-and-novel-imagery-providers/"><em>https://www.arturo.ai</em></a><em> on December 16, 2020.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e5ac26f870e1" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Growing from Start-up to Scale-up: The Best Ways You Can Support Your Employees — Arturo]]></title>
            <link>https://arturo-ai.medium.com/growing-from-start-up-to-scale-up-the-best-ways-you-can-support-your-employees-arturo-1a965fdad193?source=rss-d736cf5bbc00------2</link>
            <guid isPermaLink="false">https://medium.com/p/1a965fdad193</guid>
            <category><![CDATA[scaleup]]></category>
            <category><![CDATA[startup]]></category>
            <category><![CDATA[employee-engagement]]></category>
            <dc:creator><![CDATA[ARTURO]]></dc:creator>
            <pubDate>Fri, 11 Dec 2020 16:51:28 GMT</pubDate>
            <atom:updated>2021-03-08T20:15:19.517Z</atom:updated>
            <content:encoded><![CDATA[<h3>Growing from Start-up to Scale-up: The Best Ways You Can Support Your Employees — Arturo</h3><p>As companies grow from start-ups to scale-ups, a wave of change is felt throughout the organization. At this transitional point, you most likely have aggressive revenue targets. This means you’re working diligently to acquire more leads, close more deals, evolve your product, and offer a great experience for your new clients.</p><p>Growing your team, beyond the start-up founders and early employees, is key to meeting your business goals. It involves hiring marketers to generate leads, account executives to drive sales, product managers to improve the product, and customer success managers to support your customers.</p><p>Adding these new team members, which may increase your employee count from 15 to 30 in just a few short months, can be a challenge for any organization. How do you continue to enable all employees while maintaining the culture and vision of the company?</p><p>In this article, I share how companies can successfully support all employees as they transition from start-up to scale-up.</p><h3><strong>Defining the Start-Up from a Human Capital Perspective:</strong></h3><p>Start-up companies typically consist of 10 to 20 early employees. This group of people may have similar skill sets, and everyone on the team is actively driving toward one shared goal while wearing many hats.</p><p>At this stage, it is common for your team to have unconscious bias when making new hire decisions. Everyone holds unconscious beliefs about various social identity groups; these biases stem from one’s tendency to organize social worlds by categorizing. This means executives and founders may be in a pattern of hiring people who have parallel skill sets or professional backgrounds. Start-ups are scrappy. The core team is committed and used to “doing whatever it takes” to accomplish the shared mission. Most start-up employees are considered generalists rather than specialists.[.1]</p><h3><strong>Defining the Scale-Up from a Human Capital Perspective:</strong></h3><p>Transitioning from start-up to scale-up looks like increasing your staff from around 10 to 20 employees to 30 to 50 employees over the course of a few months. At this stage, your hiring decisions revolve around your business goals. You may be bringing in folks to manage the back office and adding in roles across business development, sales, marketing, and customer success[.2] .</p><p>Your organization may start to build distinct swimlanes and remove responsibilities that overlap. This could mean some folks need to let go of certain responsibilities and trust that leadership is hiring the best talent to run their respective roles and responsibilities. This isn’t always easy, and it requires role clarity.</p><p>The roles you hire for are more specialized at this point. To attract the best candidates for these new specialized roles, your People Team needs to develop a repeatable infrastructure to hire candidates from diverse backgrounds and experiences. It’s important to diversify your workforce while maintaining an inclusive working environment.</p><h3><strong>Best Practices to Support Your Employees as You Scale Up:</strong></h3><p><strong>New Hire Processes<br></strong>As a Chief People Officer, I recommend setting up a few key processes to ensure smooth hiring, onboarding, and continuous employee enablement. These processes include the following:</p><ul><li>Tight Interview Process — Create timelines, make sure your hiring managers know what they’re responsible for, and communicate with all parties regularly (including candidates).</li><li>Bonus and Compensation Structures — Define what this looks like; share this with your employees from the start.</li><li>Education and Training — Start-ups run so fast and so hard at the beginning that you may not consider how you onboard new hires. Think about onboarding education and ongoing training for all employees and how to educate your current team; inform them about why and how these changes may affect them.</li><li>One-on-One Meetings — Formal or informal, creating a process for managers to continuously meet with employees is crucial.</li><li>Performance Reviews — These can happen quarterly or bi-annually. Either cadence is sufficient, just make sure you are consistent; they should not be a surprise to your employees. Employees should have an understanding of how they are doing at all times through constant communication and feedback loops.</li></ul><p><strong>Career Progression</strong><br>As businesses scale up, some early employees may experience people getting hired in roles above them or more senior to them. Previously, these employees may have run a function independently and now have to navigate a new manager. To support these scenarios, People Organizations need to think about the following:</p><ul><li>Working with each functional leader to develop career paths.</li><li>Setting up clear career progression plans.</li><li>Defining clear metrics on how employees can get from A to B.</li><li>Creating internal training and professional development opportunities.</li></ul><p><strong>Build Out Management Competencies<br></strong> Start-ups promote very good doers. As you scale up, your company needs to support folks in management roles by building in management competencies. This involves training them on how to manage other team members and how to progress beyond an individual contributor role. This does not happen overnight; it’s a commitment and an investment in your employees.</p><p><strong>Motivate and Incentivize Employees</strong></p><ul><li>Offer equity options.</li><li>Offer bonus and compensation plans and share the “why” behind those plans with the team. Communicate the market value.</li><li>Offer consistent reinforcement behind the vision and explain how it all ties to the end game.</li></ul><p><strong>Get Employee Feedback, and Do Something with It<br></strong> As the Chief People Officer at Arturo, one of my first initiatives in my first three weeks was to have a 1:1 with everyone in the biz: What’s going well? Where are you being challenged? What do you expect from the people function? I then distilled everything down, looked at all the responses, and if 75% or more of the population acknowledged a concern, I addressed that with the team: We hear you. We appreciate the feedback. This is how we’re responding. Here’s the timeline of when we’ll address the concerns. Communicating this way helps manage expectations. Here are some other ideas for collecting feedback</p><ul><li>30, 60, 90 Day Individual Plans — These individual plans align with the team goals. This helps leadership learn about any barriers. It also requires constant two-way communication with your employees to let them know what you’re doing with their feedback.</li><li>Informal Check-Ins — These are outside of regularly planned 1:1s and can be especially helpful for new team members. Find out how new employees are feeling, and make sure they feel comfortable and supported.</li><li>Set Up a Mentor Program — This is an opportunity for employees to meet cross-functionally in an inviting way that creates an environment for feedback. Getting this kind of feedback is why setting up mentors is really helpful. As Chief People Officer, I meet with the mentors at the 30-, 60-, and 90-day mark to intake their feedback.</li><li>Surveys — Whenever there’s a company-wide change, send a survey to all employees to help you understand how the changes are impacting them.</li><li>Create a Buddy System — Pair new hires with a colleague to help ensure everyone is up to speed and understands how the onboarding process is working.</li></ul><p><strong>Leadership Training<br></strong> As your company grows, the people you have in leadership roles may not be fully equipped to handle all the changes from new hires to new processes to new communication channels. Put simply, leadership training and education may look something like, “This isn’t how you’ve always done it.” Or, “The way you got here may not be the way to get you there.”From a leadership point of view, conducting business and managing your employees needs to shift because the size of the team is much larger. One big change is how companies communicate and distribute information. Clear training and guidelines from the top down will help keep communication clear and expectations in check.</p><h3><strong>Final Thoughts</strong></h3><p>The transitional phase of growing from start-up to scale-up is an exciting experience for any company; however, it can come with challenges. One tough thing to do is maintain the culture you originally set in the early days with just 10 employees. Adding just three people to a 20-person team can wildly change the dynamic of an organization. Success at this phase requires that you hit your business goals and support your employees. To find success, try out a few of the best practices I’ve provided. At Arturo, we are committed to enabling our team as we grow. If you’re interested in becoming a part of the awesomeness that we’re building, check out our <a href="https://jobs.arturo.ai/jobs/Careers">open positions here</a>.</p><p><em>Originally published at </em><a href="https://www.arturo.ai/growing-from-start-up-to-scale-up-the-best-ways-you-can-support-your-employees/"><em>https://www.arturo.ai</em></a><em> on December 11, 2020.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1a965fdad193" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How Partnerships Drive Innovation — Arturo]]></title>
            <link>https://arturo-ai.medium.com/how-partnerships-drive-innovation-arturo-7a3031b5feca?source=rss-d736cf5bbc00------2</link>
            <guid isPermaLink="false">https://medium.com/p/7a3031b5feca</guid>
            <category><![CDATA[satellite-imagery]]></category>
            <category><![CDATA[geospatial-imagery]]></category>
            <category><![CDATA[imagery]]></category>
            <category><![CDATA[innovation]]></category>
            <dc:creator><![CDATA[ARTURO]]></dc:creator>
            <pubDate>Tue, 17 Nov 2020 08:00:51 GMT</pubDate>
            <atom:updated>2021-03-08T20:07:43.125Z</atom:updated>
            <content:encoded><![CDATA[<h3>How Partnerships Drive Innovation — Arturo</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/480/0*vbZSdVhieeMl0Xql.jpg" /></figure><p>Insurance carriers recognize the importance of imagery as a crucial input to meet their data requirements. However, determining the best imagery to use and how to integrate it into their systems and internal workflows can be very challenging. Insurance companies typically operate using annual or quarterly imagery, which can limit the understanding of risk due to changes at the property after image acquisition. This lack of temporal property information poses a real challenge for underwriters when it comes to accurately valuing properties and making critical risk decisions.</p><p>The global geospatial imagery analytics market is expected to grow from <a href="https://www.prnewswire.com/news-releases/the-global-geospatial-imagery-analytics-market-is-expected-to-grow-from-usd-6-9-billion-in-2020-to-usd-27-9-billion-in-2025--at-a-compound-annual-growth-rate-cagr-of-32-1-301071837.html">$6.9 billion in 2020 to $27.9 billion in 2025 </a>, at a compound annual growth rate (CAGR) of 32.1% during the forecast period. A major growth factor is the increasing demand for location-based applications, including demand from insurance carriers. Taking a client-centric approach to solving challenges across the insurance industry is at the heart of everything we do at Arturo. To better enable insurance companies with property analytics via multi-source imagery, Arturo is constantly seeking today’s most cutting-edge providers across data, technology, and imagery. We strive to build partnerships with the ‘best of the best’, particularly during these uncertain times as we remotely manage the nuances of the property insurance life cycle during an international pandemic.</p><p>In this article, we share the types of partners we seek and the three reasons why we focus on partnerships to solve the insurance industry’s most pressing challenges.</p><h3>The Types of Providers Arturo Partners With</h3><p>At Arturo, our partnerships are a key differentiator. Through partnerships with best-in-class providers, we offer our insurance carrier customers the most progressive technology in the aerial imagery space. Our team is always on the lookout for additional imagery sources, strategic partnerships, and innovative technology to enhance our process and push the boundaries on what we can deliver to our clients.</p><p>To deliver the most accurate and current property insights, we leverage multi-source and temporal imagery including aerial, satellite, stratospheric balloons, and ground inspection imagery. The partnerships we build reflect our desire to deliver the most complete and comprehensive property coverage possible.</p><p>We are proud to welcome and have these leading data providers as partners:</p><h4>Stratospheric Balloons</h4><p><strong><em>Urban Sky</em></strong></p><p>Via the new stratospheric Microballoon™, Urban Sky delivers broad area, high-resolution aerial imagery at a cost drastically less than satellites, aircraft or drones. The company’s novel invention, the high-altitude Microballoon, leverages advances in component miniaturization and first-of-its-kind material re-usability to exponentially reduce the cost of high-resolution remote sensing and weather-related data capture over broad areas, leading to more impactful data over more areas than ever before.</p><p><strong><em>Near Space Labs</em></strong></p><p>As a stratospheric imagery provider, Near Space Labs have added the production high resolution, high frequency imagery to its collection of overhead imagery for clients. The partnership with Near Space Labs increases the collection of aerial imagery for properties from two to three times per year to monthly, or even daily in the event of a natural disaster such as flooding, wildfires or coastal storm surge.</p><p>My colleague, Dr. Ben Tuttle, Ph.D., CTO of Arturo, shares our excitement, “Near Space Labs with their high-frequency observations and ability to include precise resolution is a great addition to our services. When a region or state is threatened by a natural disaster, we need to monitor properties in near real-time. Near Space Labs will help us be more responsive to our clients, whenever they need us.”</p><h4>Satellite</h4><p><strong><em>Maxar Technologies</em></strong></p><p>A trusted partner and innovator in Earth intelligence and space infrastructure, Maxar’s unique approach combines decades of deep mission understanding and a proven foundation of commercial technology to deliver solutions with unrivaled speed, scale, and cost effectiveness.</p><h4>Aerial</h4><p><strong><em>Nearmap</em></strong></p><p>Nearmap provides high resolution ortho and oblique imagery products across 70–75% of the US population, in addition to providing up-to-date and frequent coverage in Canada, Australia, and New Zealand. Accessing frequently captured aerial contents gives insurance providers the insights they need to make critical decisions across the entire policy lifecycle, from underwriting to renewal to claims and post-catastrophic events. With its recently introduced Grey Sky program, Nearmap is also helping manage post-catastrophic events.</p><p><strong><em>Vexcel Imaging/GIC/Geomni</em></strong></p><p>This partner offers state-of-the-art digital sensor systems and software solutions based on the latest and most-advanced technology. The comprehensive aerial camera portfolio provides a wide range of imaging capabilities from wide-area mapping (UltraCam Condor) to nadir (UltraCam Eagle &amp; UltraCam Falcon) and oblique (UltraCam Osprey product line) camera systems. Vexcel Imaging provides high resolution blue and grey sky imagery across more than 76% of the US population with ortho and oblique coverage, in addition to providing post event imagery after a major peril event (‘grey sky’). Their GIC program seeks to capture approx. 500,000 sq.km per annum.</p><p><strong><em>Aerometrex</em></strong></p><p>An established, successful and professional aerial imaging and mapping company, with specialist expertise in photogrammetry, LiDAR, and 3D modelling, Aerometrex has built a very strong national and international reputation as a leading practitioner of these services. Aerometrex is known and trusted by government departments at all levels as well as major private sector corporations across a range of industries including insurance.</p><p>We have also partnered with industry leaders American Family Insurance, Light Box, Nvidia, AWS, and Labelbox.</p><h3>Why Partnerships Matter</h3><p>To better service insurance carriers and enhance our technology-on-demand, deep learning models that derive accurate and current property characteristics via configurable APIs-we strive to create partnerships with the most innovative players in the market.</p><p>We have come to recognize three core reasons why partnerships are important to our mission:</p><h4>1. Realize Advanced Technology Capabilities</h4><p>The real power of our partnerships is that now, insurance carriers can take advantage of the latest technology from providers that specialize in a particular field, such as satellite imagery, as it relates to valuing and pricing properties. The unique combination of utilizing multi-source imagery, coupled with Arturo’s unique application of artificial intelligence (AI) and machine learning, delivers the most trusted data available for insurance carriers. Now, from a single provider, our customers get the most accurate property analytics to make intelligent decisions, from quoting to claims to renewals.</p><p>When it comes to our technology, Arturo’s proprietary insights are the result of our unique combination of feedback loops that continuously improve data models, access to American Family Insurance’s extensive history of claims and inspection data, and new analytics from our multi-source imagery partners.</p><p>Through our partnerships, insurance carriers can reap the benefits of technology advancements across a number of providers, all in one place. Our partnerships offer the best property insights and technology you can’t find anywhere else.</p><h4>2. Enable Efficient Investment Deployment</h4><p>To meet specific data requirements, it may be tempting for insurance carriers to develop the technology they need in-house. While this might be an admirable act, there is no need to reinvent the wheel. Recruiting the team and developing the technology is expensive, time consuming, and can be a major strain on internal resources. Plus, it isn’t core to the mission of insurance carriers.</p><p>There are already some major technology and data providers that specialize in delivering the datasets insurance carriers require to assess risk and price properties. Our list of data partners is a testament to the expertise that already exists in the market. Why not take advantage of this innovation and expertise?</p><p>Arturo’s technology, layered on top of the datasets we acquire from multi-source imagery partners, adds intelligence to enable our insurance customers to understand properties better. Now, our customers get the full picture on how commercial and residential properties have changed over time through temporal imagery or how a major weather event has impacted a portfolio with on-demand property insights. Carriers want to extract crucial information from imagery quickly and with regular currency. In other words, they want to see more “refresh”. It’s very difficult to accomplish proper coverage, accuracy, and speed, all from one provider, let alone internally. Arturo delivers “constant refresh” with temporal property insights through our partnerships. This means not only do you get the most current, up-to-date insights, but you can also select a specific timeframe against which to run an analysis. For example, you can obtain insights on a property today, before a catastrophic event a month prior and on the date you wrote the policy last year.</p><p>The partnerships we have cultivated ensure that insurance carriers are investing their money efficiently. All from a single provider, you can now unlock the power of aerial imagery with predictive analytics and proprietary insights.</p><h4>3. Develop Client-Centric Solutions</h4><p>Lastly, our partnerships provide a different lens on key customer challenges as well as an enhanced way to solve these issues. Together with these leading providers, we develop new ways to solve client challenges by focusing on each provider’s expertise for a continuous innovation. As we better understand what our clients need, we evolve our product offerings, pushing the boundaries on what is possible for the insurance industry.</p><p>As an example, one of our clients (one of Australia and New Zealand’s most recognized financial services companies) needed to select a property analytics provider that could deliver millions of property characteristics and conditions data across all urban centers and selected regional towns in Australia within forty-eight hours. We partnered with Aerometrex to provide the critical property characteristics. Our proprietary technology is designed to work with any imagery data provider, allowing Arturo to scale up infrastructure quickly to deliver the necessary data the insurance provider needed to build their pricing models. Arturo uploaded the client’s imagery, sourced from Aerometrex, into Arturo’s cloud optimized environment, aligning it with Arturo’s AI data services. Unlike any other insuretech provider, Arturo achieved a fully optimized instance to enable a fast turnaround to the client.</p><h3>Final Thoughts on Partnerships</h3><p>Arturo strives to push the insurance industry forward by taking a client-centric approach to solving challenges. By working with the best-in-class data partners, we set ourselves apart as the only provider to offer insurance carriers multi-source imagery across aerial, satellite, stratospheric, and ground. That is why we are constantly partnering with new, cutting-edge imagery providers to deliver the most complete and comprehensive property coverage possible. Partnerships enable our clients to better assess risk, identify property changes prior to renewal, and streamline quoting.</p><p>If you’re interested in becoming a partner, you can contact us by sending an email to <a href="/service/mailto:info@arturo.ai">info@arturo.ai </a>. Or, if you are an insurance carrier and are interested in Arturo’s AI property analytics solution, please <a href="https://arturo.ai/request-a-demo/">request a demo here </a>.</p><p><em>Originally published at </em><a href="http://arturo.ai/how-partnerships-drive-innovation/"><em>https://arturo.ai</em></a><em> on November 17, 2020.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=7a3031b5feca" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How COVID Has Rocked the Workplace And What You Can Do About It]]></title>
            <link>https://arturo-ai.medium.com/how-covid-has-rocked-the-workplace-and-what-you-can-do-about-it-c37465e051bc?source=rss-d736cf5bbc00------2</link>
            <guid isPermaLink="false">https://medium.com/p/c37465e051bc</guid>
            <category><![CDATA[work-from-home]]></category>
            <category><![CDATA[workplace]]></category>
            <category><![CDATA[covid]]></category>
            <category><![CDATA[work-life-balance]]></category>
            <dc:creator><![CDATA[ARTURO]]></dc:creator>
            <pubDate>Wed, 04 Nov 2020 16:37:37 GMT</pubDate>
            <atom:updated>2021-03-08T20:01:28.351Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/480/0*vADcB-ajSpVoZTcr.png" /></figure><p>When the COVID pandemic first hit the United States, many companies had to quickly adjust to full-time remote workplace environments. Leadership and People teams scrambled to figure out the best way to enable productivity, motivate employees, and boost morale during uncertain times. Some of these efforts included setting up premium Zoom accounts for everyone, shipping out large monitors to employees, and hosting virtual weekly happy hours.</p><p>Now that we are roughly eight months into remote workplace settings, many companies are pivoting to what is now the new normal. Folks in People functions, like myself, are challenged with finding creative ways to conduct work that both empowers employees every day and drives the business forward.</p><p>In this article, I unpack the challenges companies have faced during COVID, provide creative solutions to keep your employees engaged, share what we are doing at Arturo, and offer my thoughts on what is coming next in 2021.</p><p>First, let’s quickly cover some of the common challenges many workforces have experienced throughout the last eight months.</p><h3>Common Workplace Challenges During COVID</h3><p>Building meaningful relationships with colleagues is a crucial part of ensuring employees have a satisfying and rewarding workplace experience. This is especially true for new hires. Working virtually, across distributed remote offices (or homes) has made the ability to forge real relationships very challenging. This is hard under normal circumstances and COVID has magnified this challenge. And while web conferencing solutions allow employees to meet face-to-face virtually, they are not an equal substitute for in-person conversations. Keep reading this article to learn about what Arturo is doing to foster employee relationships.</p><p>When you physically come into your office everyday and have the ability to talk with your co-workers face-to-face, you can read their body language, their mannerisms, and you can get a sense of their mood. In-person working allows you to have a better understanding of someone’s reaction to an idea because you get to hear their tone. During the times of remote work, it is much more difficult to read between the lines of an email thread or long Slack thread. Some people may take tone for granted when writing an email or sending out a message.</p><p>Because your remote employees have all of the tools to work at anytime, they may feel they need to work all of the time. The lines between “working hours” and otherwise can easily become blurred when everyone is working from home. Many folks’ working schedules are different — some like to get up early and crank out a bunch of emails while others may like to respond to messages right before they go to bed. Additionally, messaging channels like Slack can often be distracting when there is a feeling to respond immediately. This can cause employees to feel like they have to work all the time. Without clear communication in place — such as working hour agreements and response expectations — your employees may feel like they need to constantly be checking in with work throughout the entire day. Later in this article, I share what we’re doing at Arturo to ensure we respect everyone’s time.</p><p>One perk that many start-up companies offer are outside-of-work events and activities like Friday happy hours and kickball leagues. During COVID, some companies are trying to replicate these social activities virtually, but unfortunately they don’t have the same impact — especially when they feel forced. And it makes sense if you think about it. Your employees have been sitting in front of a computer for the last 8, 9, or 10 hours. Do you really think they want to sit on another Zoom Happy Hour? Simply turning your happy hours virtual and hoping it sticks is not the solution.</p><p>Without in-person interaction, it may be challenging for some employees to feel motivated throughout the day in front of computers for long periods of time. Start-up offices are typically filled with the buzzing energy of teams that are all working towards a shared goal. This can be especially challenging for extroverts that thrive in team environments.</p><p>As companies begin to hire a distributed workforce, it’s important to consider everyone’s time zones when scheduling all-hands and team meetings. While the ability to hire more people across many geographies can be a hiring and recruiting advantage, leaders must keep in mind everyone’s working hours. This challenge is closely linked to challenge number 3, all about respecting your employees’ time before, during, and after working hours.</p><p>We can all agree that managing a remote workforce during COVID has not been easy. But, it is not all done and gloom. Below are a few creative ideas you can implement at your company to keep your employees engaged.</p><h3>How to Keep Your Employees Motivated During COVID</h3><p>Some companies are looking into how to handle work life balance to truly respect everyone’s time while working remotely. Some creative ideas include:</p><ul><li>Flexible schedules for parents</li><li>6 day schedules @ 6 hrs/day</li><li>4 day schedules @ 10 hrs/day</li><li>Meeting free afternoons or day</li></ul><p>At Arturo, we encourage no less than 10 hours of focus time on employees’ weekly calendars.</p><p>Communication, communication, communication! Let’s say you are an early bird that enjoys plowing through your inbox at 7am or you are a night owl that sends out emails at 11pm. We have all worked with both types of these people and at times we have felt the pressure to reply at 7am or 11pm. Communicate your working style with your employees and level set on your expectations of them. Do you want them to respond at 11pm? Most likely not. Leaders and managers should communicate working styles and response expectations with your employees. It’s important to distinguish which channels require an immediate response (i.e. Slack vs a phone call).</p><p>This is a big one. First, take a hard look at all recurring meetings and ask yourself the following questions: Are they necessary to have every week or month? What is the goal of this meeting and is that getting accomplished every time we meet? Is it imperative that everyone on the invite list is present? Do they have an opportunity to contribute?</p><p>Next, implement meeting agreements company-wide. Each meeting of more than two people should have an agenda, each invitee should understand why they are invited, stay within the allotted time, and provide opportunities for a collaborative discussion if the meeting calls for interaction. Within the meeting agenda, identify if the meeting’s purpose is to make a decision or to have a decision. Also, prereads (within the meeting description) are important for meeting participants to show up well-informed and ensure a good use of everyone’s time.</p><p>This is a new strategy we’re trying out at Arturo to foster employee relationships beyond functional teams. These cohorts can be set up in a randomized fashion. Or, you can group new hires into a cohort so they get to know other employees across departments during the onboarding process.</p><p>At Arturo we have also set up a mentoring program for all employees. Every new employee is assigned a mentor. Employees are paired together and have the flexibility of setting up how often and in which format they will meet. It’s a great way for employees to build bonds and professional development outside of the manager-individual contributor relationship. We have also organized various committees, such as health and wellness, to provide opportunities for employees to get to know people outside of their team and work on shared projects. Participating in a committee is completely optional, and people who care and want to drive a certain initiative forward will sign up. If these are mandated, the effectiveness and outcomes will drive low team engagement.</p><p>At Arturo, we set up every employee with a DoorDash account for them to use towards lunches. Employees are encouraged to take a break from work, enjoy a meal on the company, and converse with another colleague on their schedule. Services like DoorDash, Uber Eats or Grubhub make it easy and convenient for your employees to enjoy what was formally an in-person perk.</p><p>As more companies are setting up for long-term remote working, they are not renewing their office leases. This frees up funds that can be redistributed to supporting your employees mental and physical well being. Establish rewards programs around health and wellness to encourage everyone to take the time they need to care for themselves. Or apply the freed up funds to professional development and training programs. The better your employees feel, the more productive they can be at work.</p><p>These last eight months have been trying times for all of us. We all have our own personal lives, families, and adversities. With that being said, it’s crucial to treat your employees like people rather than just a resource. Think about your workforce as humans versus a business output. I advise leaders to take these action steps that we’re employing at Arturo: Talk to your employees. Get their feedback on what is working and what isn’t. Ask if they feel supported, motivated, and excited about the work they do everyday. Listen to their feedback (this is key). Tell them what you’re going to do with their feedback. Make a plan and act on that plan. Improve and iterate.</p><h3>Looking Ahead to 2021</h3><p>As companies continue to adjust to the new normal, here are a few thoughts on what we will see throughout the rest of this year and into 2021:</p><p>Many companies see that they can effectively run their businesses with remote employees and no central office. This opens opportunities to hire the best talent regardless of where they live, across the country or possibly international boundaries.</p><p>As people are staying home more often, many want more space — a bigger house, a home office, a yard, etc. During COVID we have seen many people move out of major cities and into less populated areas. Some companies are onboard with employees moving and in fact may provide breaks on relocation. Another option is to open a smaller headquarters office space, in a place where a majority of employees currently live, and people can choose to come into the office depending on how they feel. This is something Arturo has begun to explore as we continue to grow.</p><p>I am seeing more and more leaders want to learn about what other companies are doing to empower their employees. This exchange of ideas and information is happening across company and industry lines like never before. Employers are more eager to both learn how to continue to improve and share what they have found to be successful. It’s very encouraging to see as we move into this next year of the new normal.</p><p>We’re working diligently to keep the Arturo team motivated, enthusiastic and proud of the work we accomplish. And we are growing! If you are interested in becoming a part of the Arturo team, <a href="https://jobs.arturo.ai/jobs/Careers">check out our open positions today</a>.</p><p><em>Originally published at </em><a href="https://www.arturo.ai/how-covid-has-rocked-the-workplace-and-what-you-can-do-about-it/"><em>https://www.arturo.ai</em></a><em> on November 4, 2020.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c37465e051bc" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Arturo Featured on FNO: InsureTech Podcast]]></title>
            <link>https://arturo-ai.medium.com/arturo-featured-on-fno-insuretech-podcast-5b6e6e9fb158?source=rss-d736cf5bbc00------2</link>
            <guid isPermaLink="false">https://medium.com/p/5b6e6e9fb158</guid>
            <category><![CDATA[insuretech]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[podcast]]></category>
            <category><![CDATA[insurtech]]></category>
            <category><![CDATA[insurance]]></category>
            <dc:creator><![CDATA[ARTURO]]></dc:creator>
            <pubDate>Mon, 15 Jun 2020 19:48:39 GMT</pubDate>
            <atom:updated>2020-06-15T19:51:06.452Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*b2jM_WR-MqMs3Sm2MdnhTw.jpeg" /></figure><p>Recently, I had the chance to sit down with Rob and Lee at the <a href="https://www.fnoinsuretech.com/">InsureTech Podcast</a>. It turned out to be three, relatively nerdy-tech guys, talking about data science and applied machine learning. Although we were wearing headsets and using microphones, I quickly forgot the equipment and the<br>three of us fell effortlessly into the conversation.</p><p>They wanted to know how Arturo started, what we are up to now and how we are helping insurers understand property characteristics by using tools like Arturo’s Confidence Score. We talked about all the things I expected<br>to discuss and, frankly, some of the things I rarely disclose about myself.</p><p>The questions I expected were about Arturo’s spin out from American Family Insurance and one of the most popular question I get, how did we find our name — Arturo (seriously, I get this question a lot!). They asked me about insurance trends I am seeing and they told me about trends they are witnessing. We talked about data scientists (we rely on those experts a lot of at Arturo). We all admitted that we throw around that term “data scientists” and we each had a different take on what it means. How are data scientists applying their trade to our work at Arturo? Simple. They are looking for anomalies, consistencies and inconsistencies provided by our imagery and mapping to tell a story about the property — physical characteristics, along with condition, quality, and risk factors. Insurance companies hate surprises, so we try to minimize surprises about the properties in their portfolios.</p><p>When issuing a policy, the insurance company wants to know as much as possible about the property so they can accurately assess the attributes to evaluate the structure and its surrounding area. Arturo data scientists do that for them by building deep learning models. In just a few seconds, we can review a property, layer science into the process (such as artificial intelligence to observe changes in the property over time) which our customers use to better evaluate risk. Our data scientists create models that generate details for each property for our clients.</p><p>Something I didn’t expect to discuss with Rob and Lee was my passion for music and my experience composing and performing. They even incorporated a bit of my original music into the show’s intro. I realize that music helped shape me and I continue to use it as a creative outlet whenever I can.</p><p>The guys at FNO are easy to talk to and our time together flew by. They routinely interview leaders from insuretech and insurance companies to get unique and intriguing perspectives on how technology is disrupting and transforming the industry. I am very proud of what we do at Arturo and enjoy the opportunity to share our story — in this <a href="https://www.fnoinsuretech.com/podcast/episode/30d820b1/ep-78-arturos-ceo-john-isaac-jc-clark">podcast</a>, I got the chance to do just that.</p><p>John Isaac “jC” Clark | Chief Executive Officer at ARTURO</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=5b6e6e9fb158" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Getting ready for STAC 1.0]]></title>
            <link>https://arturo-ai.medium.com/getting-ready-for-stac-1-0-c260a53a42c4?source=rss-d736cf5bbc00------2</link>
            <guid isPermaLink="false">https://medium.com/p/c260a53a42c4</guid>
            <category><![CDATA[stac]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[gis]]></category>
            <category><![CDATA[open-source-software]]></category>
            <category><![CDATA[spatio-temporal]]></category>
            <dc:creator><![CDATA[ARTURO]]></dc:creator>
            <pubDate>Wed, 20 May 2020 15:18:48 GMT</pubDate>
            <atom:updated>2020-05-20T15:18:48.166Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2A5M5r7QeNXi4hhqAfRQdw.png" /></figure><p>Here at Arturo.ai we are big supporters of the Open Source Software (OSS) community. While I currently serve as Chief Technology Officer of Arturo, I am proud to have financially supported a number of OSS efforts during my previous role at the National Geospatial-Intelligence Agency (NGA). Additionally, I had the great opportunity to participate in the <a href="https://foss4g.org/">Free and Open Source Software for Geospatial (FOSS4G)</a> community and conferences, to serve on the original planning committee for WhereCamp 5280, and to act as a Charter Member of <a href="https://www.osgeo.org/">OSGeo</a>. Participating in, supporting, and gaining new friends and colleagues within the OSS community has been one of the great highlights of my career. As I say all this, I should share that the Arturo team and I are big supporters/users of Open Source Standards too. Standards like those from the <a href="https://www.ogc.org/">Open Geospatial Consortium (OGC)</a> and other de facto standards from the OSS and geospatial community have been critical building blocks of solutions I’ve built throughout my career and continue to underpin our work at Arturo.</p><p>Over the last few years one of the most exciting standards I’ve seen emerge is the <a href="https://medium.com/radiant-earth-insights/announcing-the-spatiotemporal-asset-catalog-stac-specification-1db58820b9cf">Spatio-Temporal Asset Catalog (STAC)</a>. STAC is a standard aimed at common metadata and API mechanics to search and access geospatial data. If you ever had to spend a significant amount of time trying to gather geospatial data from multiple sources and time periods you’ll understand this as an amazing step forward, too. As I’ve watched the discussion around STAC evolve it became clear to me that this would be part of our next generation geospatial services at Arturo. We are all thrilled about a truly cloud-native geospatial solution and our team has worked diligently to implement tools that will allow us to realize the benefits of STAC across our architecture.</p><p>In 2019, we had the good fortune of hiring <a href="https://twitter.com/geospatialjeff">@geospatialjeff</a> and he immediately hit the ground running with our team. Prior to joining Arturo he had already been very involved in the STAC community. You can take a look at his <a href="https://github.com/geospatial-jeff">GitHub account</a> to see some of the projects he started to help grow and implement STAC. Upon joining the company he helped our team quickly start exploring the ways that various geospatial standards like STAC, COG, “OGC API — Features”, and “OGC API — Tiles” could be leveraged to advance our geospatial architecture and services. Our team began exploring the myriad of geospatial OSS projects that would be able to help us (the list is too long to go into here). We also found that there were opportunities to build OSS solutions that would further our needs and allow us to contribute back to the community. I am thrilled to again be able to contribute back to the OSS community, the open standards community, and especially the geospatial community here at Arturo.</p><p>Our preliminary testing and implementation has shown great promise. We have been exploring STAC internally and have seen substantial simplification of our geospatial services. This has provided significant benefits to our ability to quickly create and consume geospatial data within our microservices. We have also explored STAC with some of our data partners and the results have been tremendous. We identified significant performance improvements over previous geospatial web service standards. The ease of data discovery is also an immense improvement. On-demand data discovery and ingest from multiple sources are key building blocks of our architecture so this is an exciting advancement. We firmly believe that STAC will continue to revolutionize how we discover and access geospatial data from an ever increasing range of sources.</p><p>We’ve been driving in this direction since last summer and are now ready to begin sharing what we’ve been working on. We will release a number of OSS repositories over the course of the year, each one building upon the previous. The first project we will release is <a href="https://github.com/arturo-ai/stac-pydantic">stac-pydantic</a> which appropriately contains <a href="https://pydantic-docs.helpmanual.io/">pydantic</a> data models for the STAC spec and provides a declarative and customizable interface for validating catalogs, collections, and items with python type annotations. Next up is a PostGIS-backed, modern, STAC-compliant API built with <a href="https://fastapi.tiangolo.com/">FastAPI</a> and sqlalchemy. And, last but not least, we will launch an asynchronous tile server for <a href="https://www.cogeo.org/">Cloud Optimized GeoTiffs</a> (COG) specifically designed to dynamically render map tiles from a STAC API through <a href="https://github.com/opengeospatial/OGC-API-Tiles">“OGC API — Tiles”</a>. These are all projects that we use internally at Arturo to enhance our abilities and we hope they benefit the community as well. Going forward we will share all updates and enhancements to these repositories in <a href="https://github.com/arturo-ai">GitHub</a> in hopes of helping others take advantage of STAC as well.</p><p>Ben Tuttle, PhD | Chief Technology Officer at <a href="http://www.arturo.ai/">ARTURO</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c260a53a42c4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Getting to the Heart of On-Demand Processing]]></title>
            <link>https://arturo-ai.medium.com/getting-to-the-heart-of-on-demand-processing-8236f889bde2?source=rss-d736cf5bbc00------2</link>
            <guid isPermaLink="false">https://medium.com/p/8236f889bde2</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[ai]]></category>
            <category><![CDATA[insurtech]]></category>
            <category><![CDATA[property]]></category>
            <category><![CDATA[data]]></category>
            <dc:creator><![CDATA[ARTURO]]></dc:creator>
            <pubDate>Thu, 21 Nov 2019 16:36:01 GMT</pubDate>
            <atom:updated>2019-11-21T16:36:01.486Z</atom:updated>
            <content:encoded><![CDATA[<p>In this series of engineering-focused blog posts, members of Arturo’s engineering team will dissect the infrastructure backing Arturo’s on-demand property analytics API.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*1wXUM8mFrEI-RO_7" /><figcaption><em>A sample of Arturo’s property analytics features. Our API gives unprecedented access to the latest possible property analytics data from the most up-to-date imagery available. As our machine learning services take just seconds to make a feature prediction, we’re able to offer our entire feature set on-demand — responding on average in less than five seconds.</em></figcaption></figure><p>For this first post, I’ll discuss the three types of services that compose our service graph: <strong>Machine Learning (ML) Models</strong>, <strong>3rd-Party Data Providers</strong>, and <strong>Feature Aggregators</strong>.</p><p><strong>Layers of a solution</strong></p><figure><img alt="" src="https://cdn-images-1.medium.com/max/591/0*6VlfqjvuOpVKsg3A" /></figure><p><em>Simplified service dependency graph. Some feature aggregators are dependent on other feature aggregators. For example, predicting fire risk requires the roof outline from another feature aggregator in order to localize the analysis on a specific roof.</em></p><h3>Top-level features</h3><p>How does Arturo’s API translate a simple text address to detailed information on the vast array of property attributes we support today?</p><p><strong>Feature Aggregators</strong> are responsible for the highest level orchestration logic — the first layer of services that expect an address as input, and return an output of some subset of features like roof material and condition or the location of pools on the property. These services gather all the necessary input and postprocessing context to deliver the feature set they’re responsible for in an easily-parsable format.</p><ul><li>The first step is a combination of address normalization and geocoding, which takes an address like `1747 grey avenue, 60201` and yields a normalized address (e.g. `1747 Grey Ave, Evanston, IL 60201, USA`) and coordinates (e.g. `lat: 42.0501356, lng: -87.702193`).</li><li>Translating address coordinates to a geo-referenced parcel is also part of the process, providing the full extent of a property.</li><li>Now we know the area or place on the globe we are looking for. We then fetch imagery from various satellite and aerial <strong>3rd-Party Data Providers</strong>, and pass them directly to our <strong>ML Models</strong>, which indicate, for example, where there are buildings in the images.</li><li>For some customers, it’s enough to return a simple building count or highlight the buildings in an image. Others require georeferenced objects or property characteristics in some standard format (e.g. <a href="https://en.wikipedia.org/wiki/GeoJSON#Example">GeoJSON</a>), which we derive from image metadata.</li></ul><h3>Gathering 3rd-party data</h3><p>We use <strong>3rd-Party Data Providers</strong> to access imagery and their derivatives (i.e., Digital Surface Models / Point Clouds) to support end user requirements at the property level. In order to support our strict on-demand SLA of less than five seconds, we use the following key performance indicators (KPIs) to assess each provider:</p><ul><li>Coverage: what percentage of our locations of interest are covered?</li><li>Latency: what’s the p50, p75, p99?</li><li>Volume: how many requests per time period is supported? How volatile can our request volume be?</li><li>Uptime: <a href="https://en.wikipedia.org/wiki/High_availability#Percentage_calculation">how many 9s</a> do they guarantee?</li></ul><p>Meeting this KPI standard is the first step in including a <strong>3rd Party Data Provider</strong> in our on-demand processing pipeline. However, we won’t necessarily call all providers in every request. Aerial imagery is typically higher resolution (as they fly closer to the ground) than satellite imagery, but satellite providers typically have significantly higher coverage. So for an address that has aerial coverage, we might not need a higher resolution image if we only want to analyze property-level features like “building count”. If we want to analyze features of a specific building (e.g. “roof material”), we do need a higher resolution image, so we’d call our aerial providers first and fallback to lower resolution satellite providers if there isn’t aerial coverage.</p><h3>Making predictions</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*cOWaBpRnz7CjJhrg" /><figcaption><em>A satellite image overlaid by a fire zone mask. Fire zones identify vegetation density within 50, 100, and 200 ft from the primary building of interest. The building is centered in blue, vegetation is highlighted in green, and fire zones are in red.</em></figcaption></figure><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*saSRKGlsjW3-Lm9i" /><figcaption><em>The same satellite image overlaid by a roof segmentation mask. Roof pixels are highlighted in white.</em></figcaption></figure><p>To predict “roof presence” as an example, we would first pull a high zoom image and transform it to be as close as possible to the shape the <strong>ML Model</strong> was trained on. We then pass the image to the <strong>ML Model</strong>, which returns a segmentation mask i.e., an image that says whether each individual pixel is a roof pixel. For customers who require geo-referenced objects, we project pixel coordinates to a geographic coordinate system (e.g., <a href="https://en.wikipedia.org/wiki/World_Geodetic_System">WGS 84</a>).</p><h3>So what?</h3><p>By harnessing a combination of efficient processing pipelines we’re able to deliver up to 50 highly accurate property characteristics and predictive indicators to our customers in under 5 seconds — 100% delivered on-demand.</p><h3>Conclusion</h3><p>For Arturo customers that rely on the latest property analytics, our on-demand API represents both a cost and time savings as they’ll no longer need to perform in-person property inspections.</p><p>Our engineering and applied machine learning teams continue to iterate all three types of services (<strong>Feature Aggregators</strong>, <strong>3rd-Party Data Providers</strong>, and <strong>ML Models</strong>) to improve our overall response time and make our <a href="https://medium.com/@arturo.ai/full-loop-deep-learning-c311fee8bd16">predictions</a> and <a href="https://medium.com/@arturo.ai/the-value-of-confidence-f44a9fbe20f4">their confidence</a> more accurate.</p><p>In a future engineering blog post, we’ll discuss the underlying infrastructure for all Arturo services, including hardware and common software frameworks we use.</p><p>Josh Trotter | Engineering Lead</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8236f889bde2" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Value of Confidence]]></title>
            <link>https://arturo-ai.medium.com/the-value-of-confidence-f44a9fbe20f4?source=rss-d736cf5bbc00------2</link>
            <guid isPermaLink="false">https://medium.com/p/f44a9fbe20f4</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[deep-learning]]></category>
            <category><![CDATA[insuretech]]></category>
            <category><![CDATA[machine-learning]]></category>
            <dc:creator><![CDATA[ARTURO]]></dc:creator>
            <pubDate>Thu, 18 Jul 2019 18:50:15 GMT</pubDate>
            <atom:updated>2019-07-18T18:52:38.222Z</atom:updated>
            <content:encoded><![CDATA[<h3>Introduction</h3><p>Business decisions depend more on data than ever before, yet there is no way to know the reliability of each data point used in a decision. The best option for determining the reliability of data is to validate a subset of a data source against some (often arduously) collected ground truth. Not only is this process time consuming (in some cases impossible), but it is also ineffective due to data drift (the data used to gather the model performance statistics is somehow different than the data that is used for the decision making process).</p><p>Furthermore, this process only gives you aggregate statistics on data quality. Since many decisions are made based upon one datapoint at a time, the usefulness of the data is significantly hindered (more on this later). This begs the question. How can we know the quality of each <em>individual</em> data point?</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/792/0*lLY9O5HzlnGHKTBo" /></figure><h3>Confidence</h3><p>The challenge of knowing when to trust your data isn’t new. This concept is covered at length in any publication on machine learning in healthcare. Data reliability in healthcare is imperative as diagnoses and treatments of patients based upon incorrect data could lead to serious injury or even loss of life. It could be argued that in healthcare, model interpretability and a robust measure of trust are more valuable than sheer model performance. Wouldn’t you like to know <em>why</em> your AI doctor ordered that open heart surgery for you?</p><p>In other industries there is an acceptance of the fact that a measure of trust isn’t available for most data sources, yet the data is still considered relevant and valuable. However, the unreliability factor decreases the data’s predictive power drastically. At Arturo we have conquered this unreliability factor by developing a framework for generating a measure of trust for each metric we provide. We have branded this as “confidence”.</p><h3>Arturo’s Confidence Score Framework</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*xGQ1f4BB1M0VnlQd" /></figure><p>Arturo has recognized that for our product to be most impactful and useful to our customers, we need to provide a level of trust for each data point returned. We offer confidence in one of two ways depending on the type of data we are returning:</p><h3>Probability (Discrete / Classification Data)</h3><p>This is reported as a single number between 0 and 99 that represents the probability that the classification is correct. This is only available for models that produce discrete classification data (one of some fixed numbers of choices was selected). This isn’t a true probability as one would learn about in a stats 101 course. Instead, this is an estimate of that value.</p><p>For example, given an image of a roof that we analyze and predict the roof material (a discrete class from several classes) as…</p><p>roof_material:“asphalt_shingle”</p><p>we would also return our confidence on the prediction as…</p><p>roof_material_confidence:98</p><h3>95% Confidence Interval (Continuous / Regression Data)</h3><p>This is reported as two numbers; one number represents an upper bound, and the other represents the lower bound for an interval likely to contain the ground truth. Just like the probability we provide, this confidence interval isn’t the familiar one that is taught in most undergraduate statistics course. This simply is an interval that you <em>can count on the ground truth falling within</em> most of the time (approximately 95%).</p><p>It can provide context on the direction of the error and the magnitude of the error. For example, you could tell that if we are wrong about the roof area, we are much more likely to be overestimating the area as opposed to underestimating it.</p><p>roof_area:3280</p><p>roof_area_confidence_interval:3150–3300</p><h3>The Advantage of Confidence</h3><p>Confidence can add a significant amount of value to any data used for decision making. For example, an insurance company wants to predict more accurately the risk of a home insurance policy, so it purchases data that informs them on whether or not their policyholders own pools. Let’s assume that this data was generated by aggregating local permits issued before the pool was constructed.</p><p>To get an idea of how much to trust the data, the insurance company could manually check individual properties to see how accurate the data is (there are alternatives, but the end result is the same). This would result in a single number that describes how much to trust this metric. However, there are two shortcomings to this approach. First, it is possible that this accuracy number is only correct in specific municipalities due to the different permitting laws. That is; the measure of accuracy might be incorrect when this data is applied outside those areas that were measured to assess accuracy. Second, this single estimate of accuracy does not provide the ability to know when to rely on this information. We can either trust the data all of the time, or none of the time. In most cases, relying upon the available data to determine accuracy is more advantageous than not having the data at all, however, a superior model could be developed if we could determine when to trust the data.</p><p>This problem is not specific to the insurance industry. By treating all data the same, opportunities are lost for creating new products and making better decisions. In addition to what is listed in the example above, there are other use cases where having data points that help customers understand when to trust the data are valuable.</p><p>These simple confidence indicators enable new capabilities for our customers. We have outlined a few potential advantages. but the following is not exhaustive.</p><ol><li><strong>Higher Variance Modeling</strong><br>The trust level can directly be modeled into the decision making process. When data is unreliable, then both human decision making and data driven modeling must discount this data. Knowing in which direction the model is likely to err can help determine whether or not the data should be trusted.</li><li><strong>Selective Product Quality</strong><br>In some use cases, data that doesn’t meet a certain performance level must be discarded. Using a confidence value to discard data points that are unlikely to be correct allows for products that previously would have been infeasible. <br>In the insurance industry, an insurer can choose to insure properties that other insurers might refuse, but only if the data supporting their decision is highly reliable. For example, most insurers might ordinarily refuse to cover homes in a particular area deemed high risk due to regional wildfires. However, if an insurer had access to data that correlates highly with low fire risk for specific properties in the region, then it might be able to insure these particular properties. <br>In other words, highly reliable data can provide an insurer with the confidence to cherry pick homes in this high risk region. There is a minimum data quality required before these decisions could be made. Confidence makes this possible.</li><li><strong>Minimizing Data Drift</strong><br>Generally, precision, recall, and accuracy determine the reliability of a particular model in the aggregate. However, these metrics can be highly skewed depending on the data collection process (as in the pool permit example above). A random sample of addresses may or may not match the data that was used to create the performance metrics. This could result in either trusting the data too much or too little.<br>Confidence allows the end user to get an idea of the model performance on any arbitrary dataset without needing ground truth labels.</li></ol><h3>Calculating Confidence</h3><p>How can it be possible to know how likely data is to be correct?</p><p>Have you ever looked at a picture from a long time ago and tried to identify all of the much younger faces? Usually you can identify a few people because they mostly look the same. Then using information external to the image, such as the date of the image, you reason about who you knew at that age. After using all of the available information at your disposal, a combination of the image and your memories, you can come up with a level of confidence for the identities of each person in the image. You might be absolutely certain that you recognize you and your sister, but you notice someone partially visible that you think with about 50% confidence is your best friend since she was always around.</p><p>This same reasoning is the premise behind how Arturo’s confidence modeling works. We take all the available data that we have access to — neural network parameters, image quality, prior knowledge about objects, and others — and reason about how likely we are to be correct.</p><p>For the more technical reader, we use a combination of machine learning and monte carlo simulation to estimate our confidence levels. We are in the midst of some heavy-duty research and hope to release a white paper on the inner workings of our approaches soon.</p><h3>Measuring the Quality of Confidence</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*CuVrxc4MkD8A7J8k" /></figure><p>To know if confidence is working as intended (and to know if our research is improving the confidence framework) we came up with a few robust measures of performance. For classification models we didn’t have to be so creative because we found inspiration in these articles on <a href="https://arxiv.org/pdf/1706.04599.pdf">calibration</a>, <a href="https://arxiv.org/abs/1904.01685">measuring calibration</a> and <a href="https://arxiv.org/pdf/1906.02530.pdf">trust in uncertainty</a>. These papers do a great job on measuring the quality of confidence for classification models. While we’ve drawn heavily from the principles proposed in the above research when appropriate, it is worth noting that we measure our confidence scores for classification and regression tasks differently.</p><h3>Classification Confidence</h3><p>This is measured using a metric called expected classification error (ECE) and negative log loss (NLL). ECE bins all of the predictions based on their confidence level and measures how far the average confidence is from the accuracy level of each bin. This is useful, because it is easy to inspect where the framework is failing, but it has some potential pitfalls. If every confidence prediction is equivalent to the average accuracy then the ECE will be 0. Conversely, a lower NLL will always mean that the confidence metric is better. For more information read the most recent link posted above.</p><h3>Regression Confidence</h3><p>To measure the effectiveness of confidence intervals we look at the average interval width and what percentage of our test data falls within the confidence interval. Our objective, as we work to improve this metric, is to shrink the confidence interval and still have the ground truth inside of the interval 95% of the time.</p><h3>Conclusion</h3><p>Having awareness of the uncertainty in data can be a huge advantage when making business decisions. By knowing when to trust data, how much to trust data, and how to account for the error in data, decisions can and will be much more informed.</p><p>While the entire team at Arturo is deeply excited about the science of deep learning, we’re equally if not more passionate about assuring the outputs can have high value and utility for our customers.</p><p>— Matt Sokoloff, Applied Machine Learning Engineer at <a href="https://medium.com/u/d736cf5bbc00">ARTURO</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=f44a9fbe20f4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[It takes a village…]]></title>
            <link>https://arturo-ai.medium.com/it-takes-a-village-20fabce49835?source=rss-d736cf5bbc00------2</link>
            <guid isPermaLink="false">https://medium.com/p/20fabce49835</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[insurance]]></category>
            <category><![CDATA[insurtech]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[real-estate]]></category>
            <dc:creator><![CDATA[ARTURO]]></dc:creator>
            <pubDate>Wed, 03 Jul 2019 19:56:18 GMT</pubDate>
            <atom:updated>2019-07-03T19:56:18.675Z</atom:updated>
            <content:encoded><![CDATA[<p>The saying goes, “It takes a village to raise a child,” meaning that it takes an entire community interacting with a child for that child to safely/healthily grow and succeed in the community. Suppose if you will that roughly the same is true of the AI/ML community and the resulting models that are consumed. In order to provide the actual breadth of models and information needed by the world of potential users, it will take a village of providers to create the landscape of necessary capabilities. I don’t believe any one provider is in a position to provide all the models and inferences that the world will need, but I do believe in the need for targeted specialization around specific industries and use-cases.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*JTPL_ngT0BpIuWUX" /><figcaption>Photo by <a href="https://unsplash.com/@aleskrivec?utm_source=medium&amp;utm_medium=referral">Ales Krivec</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>In that vein, I am excited for the opportunity to lead Arturo as CTO as we work to take our place in the AI/ML community and help to advance these offerings. I’ve spent most of my career building near-real-time spatial data processing and analysis systems, and this is an exciting next step in that journey. Most recently, I was working to bring automated analytics, predictive analytics, and emerging AI/ML capability to bear on the unique demands at the <a href="https://www.nga.mil/">National Geospatial-Intelligence Agency (NGA)</a>. During my time at the NGA, I was exposed to a rich array of capabilities (spatial and non-spatial) leveraging AI, ML, DL, and CV to solve unique challenges at rates and scales never before seen. When I had the opportunity to see what was being built at Arturo, the decision to join the team was a simple one. It was clear to me that Arturo had something truly unique and was decidedly in the top 10% of all the geospatial analytic capabilities I had seen and experienced over the last few years. As such, I am excited to be part of the Arturo team bringing a suite of new offerings to this burgeoning community that will add unique new value.</p><p>Here at Arturo, we hold that AI/ML capabilities will advance our understanding of the world from a macro scale all the way down to a micro scale. We are fortunate to be surrounded by amazing companies and academic/government researchers who are building excellent models for macro scale (global and country level) down to regional and local inferences and predictions. At Arturo we are focused on the very micro scale of a specific address. I like to refer to it as <strong><em>focused and targeted insights</em></strong>. We take a residential address as an input and provide inferences specifically about this address in time and space. And we don’t rely on pre-processing and caching, rather we are delivering on-demand processing on an address by address basis.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*HwDy_i9Kc07oMlTh" /><figcaption>Photo by <a href="https://unsplash.com/@britozour?utm_source=medium&amp;utm_medium=referral">Briana Tozour</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p>We have built a carefully curated set of models that answer specific questions asked by the residential P&amp;C insurance industry. We had the unique opportunity to have <a href="https://medium.com/@arturo.ai/hello-world-1c160700512d">started our work inside American Family Insurance</a>, where we received critical exposure to the needs and wants of a key customer. Using these lessons, we were able to build a solution that is keenly aimed at the specific needs of the residential P&amp;C insurance community.</p><p>Our offering was designed using four foundational building blocks: On-demand Processing , Confidence Scores (stay tuned for a blog post on this topic in the next few weeks) , Feedback Loops, and Proprietary &amp; Unique Data. Recently our applied machine learning lead wrote a <a href="https://medium.com/@arturo.ai/full-loop-deep-learning-c311fee8bd16">great post on our Feedback Loops</a>. Here I’d like to focus on our use of on-demand processing. By offering on-demand processing, we leverage the most current and highest resolution imagery available — with our most current and performant deep learning models. This increases understanding of risk to insurers and allows us to help with prefill for insurance applications, an interesting use case that we have tailored our system to support. In the case of prefill, the results need to be current and accessible within seconds to avoid delays in the application process. There are three key pieces of our system that support this: robust data partnerships, acutely tuned models, and a highly scalable processing system that can return an array of inferences in under 10 seconds.</p><p><strong>Data Partnerships</strong></p><p>Our team has strong ties to the geospatial community and a long history of using geospatial data to monitor the world. Through partnerships with the premier providers of geospatial data in the industry, we can attain data in seconds via access and discovery APIs hosted by these providers. This wealth of partnerships allows us to access the highest quality and most timely data relevant to each query of our API — specifically the highest resolution commercial satellite imagery available from our partners at <a href="http://www.maxar.com/">MAXAR</a> (formerly DigitalGlobe) and high-resolution aerial imagery collected from our partners at <a href="https://go.nearmap.com/">Nearmap</a>.</p><p><strong>Model Tuning</strong></p><p>Our applied machine learning team is continuously exploring new data, developing new models, and refining our existing models. Our Full-Loop Deep Learning™ process consistently improves our models. Our knowledge and connections to the geospatial industry result in us constantly exploring new data options, this results in a rich pipeline of new and improved models that enhance our offerings. And finally, our team of machine learning engineers tunes every model to extract features from a specific property and its neighbors in a matter of seconds. Every new/updated model goes through rigorous testing to ensure it can keep up with our on-demand inference offerings.</p><p><strong>Scalable Processing System</strong></p><p>By harnessing several tools from the suite of capabilities offered by the <a href="https://www.cncf.io/">Cloud Native Computing Foundation</a>, the <a href="https://www.osgeo.org/">OSGeo Foundation</a>, open source tools, and internally developed proprietary capabilities, we have built a processing platform that can generate a wide array of inferences in under 10 seconds. This system allows us to ingest myriad addresses through our API each second and return property insight to our end users. It is able to scale up and down seamlessly as traffic volumes increase or decrease and can handle large batch processing jobs for customers who want to monitor a list of properties. We regularly process batches of 10s to 100s of thousands of addresses for our users.</p><p><strong>On-demand Processing</strong></p><p>This allows us to support customer prefill of an online insurance application for a specific property at a pace that does not lose the attention of a potential consumer. And we are still moving forward with new and enhanced capabilities on top of this platform. Next up will be further enhancement of our offering that will monitor changes in properties quarter by quarter. Beyond that, we’ll be looking to help further improve the understanding of risk by looking at predictive models. These new models will be focused on predicting the likelihood of risk events — including a roof needing replacement in the next 1–3 years or the potential replacement cost of a given property.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*uSNfa8Chap1H0I6M" /><figcaption>Photo by <a href="https://unsplash.com/@brenoassis?utm_source=medium&amp;utm_medium=referral">Breno Assis</a> on <a href="https://unsplash.com/?utm_source=medium&amp;utm_medium=referral">Unsplash</a></figcaption></figure><p><strong>In Conclusion…</strong></p><p>It’s an exciting time to be developing in this space, and we look forward to continuing to evolve. We are excited to expand our data partnerships and to explore collaborative opportunities with our colleagues in the geospatial and deep learning community. We have a vast field of exploration ahead, but I’m excited to see us arriving at a time when we can use automated analytics, deep learning, and AI to quickly deliver insights about our world from the global scale all the way down to a specific address in record time.</p><p>Ben Tuttle, PhD | Chief Technology Officer at <a href="https://medium.com/u/d736cf5bbc00">ARTURO</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=20fabce49835" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>