<?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 ThoughtSpot on Medium]]></title>
        <description><![CDATA[Stories by ThoughtSpot on Medium]]></description>
        <link>https://medium.com/@thoughtspot?source=rss-84a96ef19c56------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*Z7p0Uqbg-Fwh7K7J.png</url>
            <title>Stories by ThoughtSpot on Medium</title>
            <link>https://medium.com/@thoughtspot?source=rss-84a96ef19c56------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 14:51:21 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@thoughtspot/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[How to Recruit from Google]]></title>
            <link>https://medium.com/@thoughtspot/how-to-recruit-from-google-77c92feffd5a?source=rss-84a96ef19c56------2</link>
            <guid isPermaLink="false">https://medium.com/p/77c92feffd5a</guid>
            <dc:creator><![CDATA[ThoughtSpot]]></dc:creator>
            <pubDate>Thu, 18 Feb 2016 21:58:43 GMT</pubDate>
            <atom:updated>2016-02-20T00:31:01.035Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1000/1*qxy3sP8gxRhrojD1N8DzGQ.png" /></figure><p>I have tremendous respect for Google when it comes to software engineering. In the last 15 years, Google has taught the world how to build distributed systems at scale, how to use commodity hardware and open source software like Linux for mission critical applications, how to write search algorithms that are uncannily smart, and the list goes on and on. A lot of these innovations have been replicated at other web companies. Facebook, LinkedIn, Twitter, and others use infrastructure software modeled after approaches created by Google.</p><p>And now these innovations are making their way into the enterprise. At my first startup, Nutanix, we took inspiration from Google File System (GFS) to create storage software that runs on the same machines that run applications. As a result, we completely eliminated the need for expensive network storage. This breakthrough made Nutanix the fastest growing enterprise tech company of the last decade.</p><p>Today at ThoughtSpot, we’re inspired by the “human scale” UX that Google’s search engine provides to billions of people worldwide. As we all know, search at scale is a very hard problem; it was natural for us to look for talent from Google to make this a reality.</p><p>Of course, this is easier said than done. It’s hard work recruiting Google engineers. Over the years, I’ve discovered a few secrets that helped me recruit over 50% of our engineering team from Google. Here they are:</p><h4>1. Understand motivations.</h4><p>Figure out what the engineer is getting out of Google. This requires understanding his or her…</p><ul><li><strong>Drive:</strong> Most Google engineers want to change the world. They believe in Google’s mission to “not be evil.” They want to be part of something bigger than themselves. If you’re trying to attract them, you have to be fundamentally driven by the same passion — making an impact. Google engineers wouldn’t want to work with someone trying to make a quick buck.</li><li><strong>Compensation:</strong> You have to know that these people are paid really well. An engineer with a few years of experience may be making over $200K at Google. And a top-rated Googler can easily make $700K to $1M a year, maybe more. You won’t be able to hire them just by saying you’ll give them stock. Money alone is not going to lure them away.</li></ul><h4>2. Master the research organization structure and projects.</h4><p>Before sales meetings, people study each other’s profiles on LinkedIn. But before you meet with an engineer from Google, you should catch yourself up on the candidate’s research and professional background. What are these engineers working on? What papers are they reading? You should be reading those papers, too. Learn to speak their language and discuss passions or tough technical problems that they are actively pursuing at work.</p><h4>3. Go to them.</h4><p>Google’s campus in Mountain View is amazing. It’s basically an extension of college, combined with a country club and fancy food court. People don’t have to leave. What does that mean? You have to go to them.</p><p>There are a lot of meet up places nearby to choose from. I decided to make the Pear Avenue Starbucks near Google my home away from home. I got comfortable with it. I probably had 200+ meetings in the first two years at that Starbucks (it helps that it’s on my way to work).</p><p>I suggest you find your own spot (there’s also an <a href="http://www.yelp.com/biz/cucina-venti-mountain-view">Italian restaurant</a> for those of you non-Starbucks people). If you want to take someone from Google, you’re going to have to be there for a while, so you might as well pick a place you like.</p><h4>4. Zero in on the talents you want.</h4><p>Google is huge. So is the diversity of talents. If you’re trying to build a consumer app, you won’t hire the Google engineers working on the search engine. There are lots of people at Google doing different things well; you have to make sure you’re hiring the right people for your company.</p><p>Once you’ve found the right teams to target, you then need to narrow down the pool even more. You want someone who is hungry and ready for the leap to the startup world. Here are a few things to look for:</p><ul><li>Someone who has been at Google for a long time and is tired of the bureaucracy that may be slowing down his or her project. Someone who is ready to get things done and move fast.</li><li>Someone craving more responsibility and ownership. Some of the teams at Google have become very big, which dilutes the amount of ownership any one individual can feel. At a startup, one person can have a much bigger impact.</li><li>Someone who wants to work with people who are not in it for the money. If they hear from an entrepreneur who is really interested in money, most Googlers won’t bat an eye. They’re the type to evaluate a company based on questions like, “can this idea make a dent in the universe”? The code they write at Google impacts billions of people; they want to understand how your technology makes someone’s life better.</li></ul><h4>5. Connect the dots.</h4><p>Most people from Google are pretty happy. Wouldn’t you be, if you had access to gourmet food all day long? The result is that most people from Google aren’t out there browsing Indeed.com or dice.com where they might stumble across your company and job postings.</p><p>Of course, having a few Google engineers on your team creates momentum to help you hire more of them. But how do you build that foundation network?</p><p>A natural starting place is LinkedIn, but many engineers don’t have a lot on their LinkedIn profiles; they might not even have a profile because they don’t really need to advertise themselves. To work around this, find someone you like who does have a profile, and then look at “People similar to _______” on the sidebar. Meet who you can in person, and network from there. One of my toughest hires took 3 introductions to find! But once you have secured a few, you’ll gain momentum and be able to recruit others.</p><h4>6. Buckle in.</h4><p>It’s not going to be easy to get these engineers. Knowing the right places to look, tactics and language to use will be key on selling the move to a startup, but you’ll need to be patient.</p><p>To convince one of our Principal Engineers to join, it required setting up meetings with investors, an early engineer from my first startup, Nutanix, and board members before he agreed to take the leap.</p><h4>7. Most importantly, do the right thing for the candidate.</h4><p>If you don’t truly believe that you can move the needle for the engineer you are trying to recruit, please don’t. Make an effort to understand his or her goals and ensure that your company has more than a fair chance of meeting those goals, whatever they might be.</p><p>CEOs wear many hats. On a given day, I’m balancing meetings with my team, reviewing product ideas and selling our product to prospects. But there is no responsibility more important nor challenging than recruiting, which is why it has been my priority since Day 0. Having the right people on board is how great products and companies are built.</p><p><em>Originally published at </em><a href="http://www.thoughtspot.com/blog/how-recruit-google"><em>www.thoughtspot.com</em></a><em> on February 18, 2016.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=77c92feffd5a" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[America’s Mass Shootings in 6 Charts]]></title>
            <link>https://medium.com/@thoughtspot/america-s-mass-shootings-in-6-charts-8e25916ae19f?source=rss-84a96ef19c56------2</link>
            <guid isPermaLink="false">https://medium.com/p/8e25916ae19f</guid>
            <category><![CDATA[guns]]></category>
            <category><![CDATA[politics]]></category>
            <dc:creator><![CDATA[ThoughtSpot]]></dc:creator>
            <pubDate>Thu, 10 Dec 2015 08:00:00 GMT</pubDate>
            <atom:updated>2015-12-14T23:17:09.111Z</atom:updated>
            <content:encoded><![CDATA[<p>Everyone is talking about the <a href="http://www.nytimes.com/2015/12/05/opinion/end-the-gun-epidemic-in-america.html?_r=1">mass shooting epidemic</a> facing America. 35 victims in San Bernardino, California. 12 in Colorado Springs, Colorado. 19 in Roseburg, Oregon. And that’s just in the last six months. The rising number of fatalities is undeniable and terrifying for most of us. Mass shootings are happening so frequently that it’s become hard to keep up with all the stories. It’s tragic and the collective angst of our nation has us all calling for change. Everyone wants to save lives — the big question is how we should go about doing it.</p><p>Gun control is a highly polarizing topic in this country. When opinions are strongest and opposing, one of the best ways to bring people together is to show them the facts and data. To that end, the team here at ThoughtSpot used our search-driven analytics technology to analyze the public data available from the <a href="http://www.shootingtracker.com/">Mass Shooting Tracker</a>. Data collection for this project began three years ago following the Newtown, CT elementary school massacre, with a “mass shooting” defined as an occasion during which four or more people are shot.</p><p>Only with a better understanding of the facts will we find a middle ground for change. Here are six charts that help visualize the rise in mass shootings over the past 3 years.</p><h4>1. We’re averaging 1 mass shooting per day</h4><p>This chart shows weekly mass shooting trends since the beginning of June. The weeks with an average of one shooting per day are highlighted in red. America now averages <strong>8.5 mass shooting incidents per week</strong>. The highest on record is <strong>18 mass shootings in one week</strong> (July 13, 2015).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*edjWGyjoghpOcYju.png" /></figure><h4>2. Fatalities are rising fast</h4><p>The number of mass shootings is on the rise and so are the number of deaths per shooting. <strong>7 out of the 10 most fatal shootings took place this year</strong>. This chart shows the top 10 most fatal mass shootings over the past three years. 14 people died in San Bernadino last week.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*GkHLL2zkI7sHtBnk.png" /></figure><h4>3. Mass shootings are everywhere</h4><p>This map plots the last three years of mass shootings in the U.S. by location. The only states without mass shootings were Hawaii, North Dakota, New Hampshire, and Wyoming. <strong>30% of shootings occurred in the same 20 metro areas</strong>.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*LPX8F9LbjzIY_lnu.png" /></figure><h4>4. Chicago has the highest number of shootings and fatalities</h4><p>Chicago has <strong>2.5 times as many incidents in the last 3 years as the next highest city</strong> (New Orleans). The size of the bubble represents the number of shootings. The height on the chart represents the number of fatalities.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*J5G_kynv0Q1nM376.png" /></figure><h4>5. Louisiana is the most dangerous state</h4><p>Even though California accounted for the most shootings (124) over the last three years, <strong>Louisiana mass shootings have nearly 3x as many victims per capita</strong>. This map shows the victims per capita (per 100,000 residents) for states with more than 5 mass shootings.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*rO-KGcpW6-WUFWL_.png" /></figure><h4>6. States with fewer gun laws have more gun deaths</h4><p>Texas, a state that is known for having few gun laws, has an <strong>average fatality rate that is 43% greater than California </strong>and 95% greater than Illinois. This tree map plots the average fatality rate (depth of color) against the number of incidents (rectangle area).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*qCYQVVlp4tFLHtAD.png" /></figure><p>We hope this data helps explain the magnitude of the mass shooting epidemic facing our nation. We can keep surfacing the data, but it falls upon all of us to keep having the tough conversations. If you found this analysis useful, please share it with your local elected officials to help shine a light on these tragedies.</p><p>And if there’s another dataset you’d like to see analyzed in ThoughtSpot, feel free to send it to marketing@thoughtspot.com.</p><p><em>Originally published at </em><a href="http://www.thoughtspot.com/blog/america%E2%80%99s-mass-shootings-6-charts"><em>www.thoughtspot.com</em></a><em> on December 10, 2015.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8e25916ae19f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How to Design Enterprise Apps for Humans]]></title>
            <link>https://medium.com/@thoughtspot/how-to-design-enterprise-apps-for-humans-630f926ede96?source=rss-84a96ef19c56------2</link>
            <guid isPermaLink="false">https://medium.com/p/630f926ede96</guid>
            <dc:creator><![CDATA[ThoughtSpot]]></dc:creator>
            <pubDate>Thu, 18 Jun 2015 21:06:50 GMT</pubDate>
            <atom:updated>2015-06-18T21:06:50.399Z</atom:updated>
            <content:encoded><![CDATA[<h4>by Jonathan Moreira</h4><p>I love that consumer technology is evolving everyday. Every time you log into your favorite apps, they’re getting better. It’s awesome. But the enterprise space tells a different story. Most of the tasks business leaders need to perform on a daily basis — analyzing billions of rows of data, creating revenue reports for managers, performing complex queries to gain business insights- are technically difficult. Yet the tools at our disposal for these tasks are outdated and clunky. You can’t Google this month’s business summary on your iPhone five minutes before your meeting.</p><p>Why is that? We should expect more from enterprise technology.</p><p>I joined ThoughtSpot in 2013 with a “beginners mind.” I wanted to build an enterprise analytics app for humans — something for the everyman, not for data analysts. I wanted to build a powerful app that would help someone do their job faster and easier, not one that would overwhelm them with sophistication. As we’ve built our product the past two years, we’ve borrowed a lot from consumer app design.</p><p>Here are my four biggest takeaways for designing human-scale enterprise apps.</p><h3>1. There’s no finish line to search.</h3><p>When we use apps today, we think, we wonder, we ask, and we expect an answer. This is the way our brain works. You type something, you get something. That’s it. It’s why having a search bar in your app is so important.</p><p>But building a good search experience — specifically, the search bar and the interactions around it — is a tricky thing to design. Let’s face it, we’ve all used a number of different search bars in products and there’s a big difference between good and great. It’s why we are always refining our search experience as we learn more about our users and the way they think.</p><p>The big a-ha moment for us came when we realized people needed help building their searches. Often people don’t know what they don’t know, or don’t know which questions to ask. Similar to how search suggestions guide much of your Google and Pinterest search experience, our recent UX studies taught us that search string readability is crucial. So we recently introduced boxed search terms around each keyword you type in the search bar, making it easier for users to parse and make sense of the query. Clarity <a href="https://medium.com/the-year-of-the-looking-glass/the-5-most-common-mistakes-in-design-52816ab77b3d">matters</a>. You can see below how we’ve borrowed from Pinterest’s search experience. So far the customer feedback has been fantastic.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*LbyftJ-IfP4IWvxy.png" /></figure><h3>2. Never sacrifice speed for frills.</h3><p>Your user experience, no matter how beautiful or intuitive means nothing without speed. Research shows that if a Google search takes more than two seconds to load, most people will abandon it for another page. But enterprise software is notorious for creating tools with excessive numbers of features, which ends up creating bloat and ultimately slowing down the product. Nothing is more important than speed. Nothing.</p><p>So we knew we needed a simple search experience, but that it also had to stay fast even as we tackle massive data scale. That’s when things got really hard. As a designer, my goal is to create solutions that are usable and visually appealing. Users don’t care what’s going on behind the curtains — they just want a usable product and clear answers as quickly as possible. But I’ve had to take more things into account with this product. For example, we recently made the decision to move the position of the headlines of charts from the top of the page to the bottom. The top placement gave users an immediate and clear understanding of the answer to their queries while looking nice visually — which was my goal — but it was slowing down the product performance. While the backend caught up, we had to do a design half-step. Sometimes we have to make sacrifices in terms of aesthetics to improve the overall user experience.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*JEXTvLq-JDEeRg3C.png" /></figure><h3>3. Simplicity rules.</h3><p>So now you’ve got speed. But if you talk to any product designer, he or she will also tell you that simplicity is key. Whitespace is celebrated all the time. Yet if you go look at most enterprise apps, they clearly missed the memo. It’s trickier than it looks. When I started at ThoughtSpot as the first product designer, I came up with something clean and simple. But then came the feature requests. And more feature requests. It’s great to have so many requests from users, but I’ve had to learn how to prioritize.</p><p>You might remember Marissa Mayer’s reign as Google’s “<a href="http://www.fastcompany.com/56804/beauty-simplicity">high priestess of simplicity</a>.” She rose to that prominence because she fought to keep Google’s homepage as simple as possible. Hundreds of people from different parts of the organization pushed back on her to add to it, but she held her ground. I think about this whenever engineers ask for certain features. I often have to say no. For example, an earlier release of our product included a data panel which gave the user options to explore available data before starting a search. While it was very helpful for some users, others didn’t ever use it, even though it was taking up a lot of real-estate by default. So we made it optional — the user can now close it and really focus on the search experience. A minimal interface matters much more when you’re trying to scale to millions of untrained users.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*ODWO6xWRynCrSgEb.png" /></figure><h3>4. Surprise and delight.</h3><p>Everyone thinks about UX in terms of clean aesthetics and a simple user experience, but there’s a third dimension to it that’s harder to spot. It’s often this dimension that will take your app beyond useful to “sticky” and “loved.” I call it the hidden gem dimension. This is where you’ll find features that are incredibly useful but not always noticeable.</p><p>For example, I just came across two hidden app features that I use daily:</p><ul><li>The “Pin” action of a Chrome tab. You can easily keep an important website always open in your tab as a default.</li><li>The option to “Create custom emoji” on Slack. Totally unnecessary, but it makes the app incredibly engaging. I love it.</li></ul><p>While not crucial to my browsing or email experience, these features are valuable. The first time user experience is powerful — when I discovered each of these, my first thought was, “Wow.” They bring real value. To incorporate this dimension into our product, we’re adding an additional collaboration feature that allows users to leave comments on data. It’s a small thing, but it makes a big difference. Leaving a note on a chart you made for the team might save you a few emails or meetings, making everyone more productive. That’s pretty cool.</p><p>I don’t want ThoughtSpot to become an overly complex BI solution with a heavy interface that users can’t understand. But design here, like at any company looking to disrupt, is still a work in progress. It took Google over a decade and an army of designers and engineers to nail the experience they have today. Our goal is not only to bring a cool app to market, but more importantly to change enterprise technology expectations. Designing an enterprise app with consumer-scale ambition gives us an opportunity to make a huge impact.</p><p><em>Originally published at </em><a href="http://www.thoughtspot.com/blog/how-design-enterprise-apps-humans"><em>www.thoughtspot.com</em></a><em> on June 10, 2015.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=630f926ede96" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[How We Engineered ThoughtSpot to Deliver Search at Scale]]></title>
            <link>https://medium.com/@thoughtspot/how-we-engineered-thoughtspot-to-deliver-search-at-scale-46a67dad91b9?source=rss-84a96ef19c56------2</link>
            <guid isPermaLink="false">https://medium.com/p/46a67dad91b9</guid>
            <dc:creator><![CDATA[ThoughtSpot]]></dc:creator>
            <pubDate>Fri, 17 Apr 2015 18:43:00 GMT</pubDate>
            <atom:updated>2015-04-17T18:43:00.390Z</atom:updated>
            <content:encoded><![CDATA[<p>by Abhishek Rai</p><p>Here at ThoughtSpot, we don’t believe business leaders should have to wait for days or weeks to get a report built by IT or data analysts. The beauty of search-based analytics done at scale is that it can dramatically reduce the time to get insights from data down from months to seconds — for anyone.</p><p>Search done well always looks so simple. However, building the technology to scale search-based analytics across billions of rows of data is anything but.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/665/1*nDG3vwwuNq0KngxR04MTzw.png" /><figcaption>Search done well always looks so simple. However, there’s a lot of magic happening under the hood of ThoughtSpot.</figcaption></figure><p>There’s a lot of magic happening under the hood of ThoughtSpot. Our analytical search engine architecture is composed of multiple services, such as search backends, a relational data cache, a BI engine, etc. Each of these services (or “microservices”) is an independent distributed system, communicating with each other over the network. From the moment a user enters a keystroke in the search bar until they see a result, a flurry of network calls are made to deliver the perfect answer to the user.</p><p>We had to make a number of complex engineering decisions to build our system from scratch. In this post, we cover the rationale that led us to using Remote Procedure Calls and we’ll give you a peek into network communication infrastructure in ThoughtSpot.</p><p><strong>**Deep dive technology alert: If you become alarmed when reading technical material, you should close this browser tab now.**</strong></p><p>The simplest software abstraction for network communication is the “socket”. It’s a no frills wrapper over an underlying network connection (e.g. TCP connection) which can be used for sending raw bytes. But real systems need more than just the ability to send and receive data. For example, they need APIs, fault tolerance, ease of use, cross-platform support, etc. These needs can be fulfilled using higher level abstractions, two common choices being Remote Procedure Calls (RPC) and HTTP based RESTful APIs.</p><p>A Remote Procedure Call (RPC) looks like a function call but is fulfilled by making a network call to a server, which is usually on a different host. Remote Procedure Calls were popularized by the <a href="http://en.wikipedia.org/wiki/Open_Network_Computing_Remote_Procedure_Call">Sun RPC</a> implementation in the 1980s, and are used by a wide range of web services and distributed systems today.</p><p><a href="http://en.wikipedia.org/wiki/Representational_state_transfer">REST</a> (REpresentational State Transfer) is an architecture style for defining network APIs. Proposed in 2000 by Roy Fielding, HTTP based RESTful APIs have become the most popular way of creating scalable web services.</p><p>In ThoughtSpot, we use both RPCs and JSON over REST heavily. RPCs are primarily used for communication between backend services. JSON over REST is exposed by the backend web service for consumption by the web frontend.</p><p>Using REST for web services offers several benefits:</p><ul><li>Self-documenting: RESTful APIs combine HTTP verbs (GET, PUT, POST, etc.) with hierarchical “resources” which represent entities or concepts within the system. The combination facilitates exploration and tends to be self-documenting.</li><li>Separation of concerns: ThoughtSpot exposes a rich set of features through the UI. The resource hierarchy in REST allows us to easily separate different feature sets by exposing them as different resources.</li></ul><p>For communication between backend services, we have adopted RPCs. Unlike JSON over REST, RPC frameworks can provide greater efficiencies in CPU, memory, and network bandwidth. They can also provide better type safety than JSON. RPCs are also a more intuitive programming model since they appear as function calls in the code.</p><p>Our internal RPC library uses <a href="https://thrift.apache.org/">Apache Thrift</a> as the transport and <a href="https://developers.google.com/protocol-buffers/">Google protocol buffers</a> as the message format. It provides additional features on top, such as name resolution, connection pooling, fault tolerance, distributed tracing, and logging. We have been using this framework in production for over a year now across 4 programming languages: C++, Java, Python, and Go.</p><p>Developers have many RPC frameworks to choose from today. Based on our experience, we think that following features are desirable in an RPC framework:</p><ul><li>Fast, event-driven communication</li><li>Efficient wire encoding and decoding</li><li>Space efficient wire coding</li><li>Language and platform independent, declarative interfaces</li><li>Integration with common programming languages</li><li>Support for synchronous and asynchronous communication models</li><li>Connection pooling</li><li>Health checks and load balancing</li><li>Authentication support or hooks</li><li>Monitoring hooks for tracking response times, queuing latency, etc</li><li>Testing hooks</li></ul><p>Recently, we started evaluating Google’s newly open-sourced, BSD-licensed RPC framework, <a href="http://www.grpc.io/">gRPC</a>. We met with the gRPC team to learn about the project and were impressed by their commitment to building a vibrant open source community around gRPC. The project is being run in the open with upstream master repositories on GitHub and a <a href="https://groups.google.com/forum/#!forum/grpc-io">Google Group</a> for discussions.</p><p>We are excited about gRPC for multiple reasons. gRPC uses HTTP2 as the transport, which comes with goodness like connection multiplexing, bidirectional streaming, flow control, and header compression. The seamless integration with protocol buffers is a great plus for us, given that we are heavy users of protocol buffers. Additionally, gRPC puts a strong focus on building low latency, highly reliable production services through framework support for health checking, monitoring, etc.</p><p>At ThoughtSpot, we’re all about delivering breakthrough performance for the end user. Designing our network communication infrastructure like this is just one way in which we’re trying to disrupt the BI industry and help our customers make smarter business decisions.</p><p>Want to see this technology in action? Join ThoughtSpot CEO, Ajeet Singh, and Principal Engineer Vijay Ganesan for a <a href="http://go.thoughtspot.com/webinar-under-the-hood-0415.html">deep-dive webinar</a> that will go “Under the hood of ThoughtSpot” on Tuesday, April 28 at 10:30am PST.</p><p><em>Originally published at </em><a href="http://www.thoughtspot.com/blog/how-we-engineered-thoughtspot-deliver-search-scale"><em>www.thoughtspot.com</em></a><em> on April 16, 2015.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=46a67dad91b9" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[9 Design Insights We Got from UX Testing on Kids]]></title>
            <link>https://medium.com/@thoughtspot/9-design-insights-we-got-from-ux-testing-on-kids-fb20a9a62195?source=rss-84a96ef19c56------2</link>
            <guid isPermaLink="false">https://medium.com/p/fb20a9a62195</guid>
            <dc:creator><![CDATA[ThoughtSpot]]></dc:creator>
            <pubDate>Fri, 20 Feb 2015 19:20:12 GMT</pubDate>
            <atom:updated>2015-02-20T19:20:12.832Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*3DXgIfrTt6kKStvtuca6-g.png" /></figure><p><a href="http://www.thoughtspot.com/blog/9-design-insights-we-got-ux-testing-kids">Alicia Avrach / November 05, 2014</a></p><p>Why test the <a href="http://www.thoughtspot.com/product">ThoughtSpot</a> user experience on children? Because we could. We had a spare pair of 9-year-old twin boys whose parents both work here. (So no children were harmed, no labor laws broken.)</p><p>You probably test your user experience in various ways: A/B testing, wire-framing, recorded stream of consciousness walkthroughs. Have you tried testing it on kids?</p><p>“But our software isn’t meant for kids,” you say.</p><p>Stick with me for a minute. I took a class at Stanford’s <a href="http://dschool.stanford.edu/">d.school</a> last Spring. One design thinking method we learned was to try testing outside the normal range of individuals you define as your “users.” This includes outliers based on geography, demographics, or use case.</p><p>Hence, the kids. But first, we had to find a data context they could understand. We chose to use historical baseball statistics from <a href="http://www.seanlahman.com/baseball-archive/statistics/">Sean Lahman</a>. Sean makes the data available for use under a Creative Commons license. (Thanks, Sean!)</p><p>Here’s what we learned.</p><ol><li><strong>Our “user” is smarter than we think.</strong> We worry a lot about the person we think our end user is. Will they know what to do? Will they get confused? Nobody reads user manuals anymore, and they shouldn’t have to. But children like to play with new things, and <a href="https://www.psychologytoday.com/blog/freedom-learn/200901/minimally-invasive-education-lessons-india">they will explore and learn on their own</a>. The kids learned how to type in their questions to get baseball stats in under five minutes.</li><li><strong>The first experience has to be good.</strong> Kids approach everything as entertainment. When you’re competing with Minecraft and YouTube for eyeballs, your stuff better be easy and fun to use.</li><li><strong>Familiar data matters. </strong>We didn’t use financial data. We used baseball stats. It’s a natural choice — since the kids love baseball cards, the data was familiar. Choosing a dataset the testers know makes it easy for them to come up with questions to ask. Remember to keep it age appropriate, too! Police blotters, for example, are interesting, but may not be the best choice for kids.</li><li><strong>Data provides a huge learning opportunity.</strong> Okay, granted: Our kids may be a bit geekier than your average fourth grader. But it is worlds more fun to explore formulas and data integrity using a search engine than with a Common Core worksheet.</li><li><strong>We are old. </strong>The world has changed. Kids are expert at search. My Google search history is filled with searches for Pokemon, Minecraft, and Angry Birds. I can’t remember when they learned how to search. It’s not like we sent them to summer camp to learn how to use Google. They learned over time on their own, by trial and error.</li><li><strong>Kids have no concept of relational data. </strong>The boys did not have a mental picture of tables in a database. They don’t know a column from a row, and SQL is just a bunch of letters to them. Yet they were able to type a few words in the search bar and work from there.</li><li><strong>Spell check is important.</strong> When you’re 9, you may have memorized the stats for the top 10 pitchers of all time, but you still type in “picher” to search for them. We needed to add a spell check feature to the product.</li><li><strong>Data is a great teacher of character. </strong>The kids searched for the pitcher with the highest number of game losses, which turns out to be Cy Young. Then they searched for hte pitcher with the most wins. If you know anything about baseball, you know that the answer is the same. The lesson? It takes a lot of tries to get a lot of success. Failure can lead to improvement if you use it to learn.</li><li><strong>Homework is all about data now.</strong> One of the kids brought home a school assignment to look up the ballpark attendance of their favorite team and chart it annually over time. The “old fashioned” way of finding the information might be to do a Google search. But he had access to our baseball database. I told him it would be find to use it, so long as he could search for the answers himself. Here’s what he found:</li></ol><p>Go on, give your kids some extra screen time. You might be surprised what you (and they) can learn.</p><p><em>Originally published at </em><a href="http://www.thoughtspot.com/blog/9-design-insights-we-got-ux-testing-kids"><em>www.thoughtspot.com</em></a><em> on November 5, 2014.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fb20a9a62195" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[2015 Predictions: BI Finally Shines a Spotlight On Everyday Business Users]]></title>
            <link>https://medium.com/@thoughtspot/2015-predictions-bi-finally-shines-a-spotlight-on-everyday-business-users-8bfc6404872d?source=rss-84a96ef19c56------2</link>
            <guid isPermaLink="false">https://medium.com/p/8bfc6404872d</guid>
            <dc:creator><![CDATA[ThoughtSpot]]></dc:creator>
            <pubDate>Thu, 19 Feb 2015 22:09:50 GMT</pubDate>
            <atom:updated>2015-02-19T22:09:50.723Z</atom:updated>
            <content:encoded><![CDATA[<p>As we ease into 2015, big changes are afoot in business intelligence. To excel, vendors will need to redirect efforts where results matter the most: the end user.</p><p>Currently, the BI landscape faces two huge obstacles: cost and complexity. Making up a $69B market, BI projects are mammoth-sized undertakings, requiring huge price tags, armies of implementation consultants, and specialized, full-time staff for ongoing support. It’s not uncommon for these projects to drag on for six months before a single report is generated.</p><p>Think about Facebook and Google. If you had to take a training course or rely on outside support to use them, would they claim billions of users? Of course not. UX simplicity is paramount.</p><p>With this focus on the end user in mind, I jotted down some thoughts on where BI is headed in 2015.</p><ol><li><strong>BI vendors will start adopting search-based interfaces with varying degree of capability. </strong>A simple, intuitive end-user experience will be an integral element of this refocus because it’s an already common and effective user experience model in the world of access to consumer information (think about how you use Facebook, Amazon, and LinkedIn). Following this model, more and more BI providers will add features that allow users to simply search across billions of rows of data to gain instant insights.</li><li><strong>Big Data adoption will remain limited because of the last-mile challenges users face in accessing data. </strong>Ignoring for now the question of how to properly define “big data,” we should expect to see platforms like Hadoop become more widely adopted to bring order to unstructured data from our daily lives (via blogs, pictures, sensors, etc.) and derive actionable intelligence from it. We’re seeing great progress in capabilities, but where is the usability? What we really need for Big Data is an easy-to-use, simple interface that gives non-technical users actionable insights.</li><li><strong>Technological advances in DRAM economics and security features will accelerate utilization of BI tools and fuel software innovation. </strong>For decades, the BI stack has been designed with hard drives in mind for storage. Now with DRAM taking the place of spinning platters, all sorts of possibilities come to life in terms of how much data can be stored in memory rather than disk and how quickly that data can be accessed. This will transform what kind of user interfaces can be built to get answers from these databases, to support more real-time, interactive and iterative user interfaces for data exploration and analysis.</li></ol><p>While a quickly-evolving big data revolution ripples across multiple industries, we remain using business intelligence technology from two decades ago. A focused shift back to the BI end user via search will drive big changes in 2015, which is a promising step forward. As is almost always the case, the simplest experience is the best one.</p><p><em>Adapted from the </em><a href="http://www.virtual-strategy.com/2015/02/09/executive-viewpoint-2015-prediction-thoughtspot-bi-2015-not-about-visualizations-its-abou?page=0,0#axzz3ReG6ubHR"><em>original article</em></a><em> published in Virtual-Strategy Magazine.</em></p><p><em>Originally published at </em><a href="http://www.thoughtspot.com/blog/2015-predictions-bi-finally-shines-spotlight-everyday-business-users"><em>www.thoughtspot.com</em></a><em> on February 18, 2015.</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=8bfc6404872d" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>