<?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 Rouz Jazayeri on Medium]]></title>
        <description><![CDATA[Stories by Rouz Jazayeri on Medium]]></description>
        <link>https://medium.com/@rouz_85704?source=rss-6a004c556e2d------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/0*NDZOXKNoMjNlTBvJ.jpg</url>
            <title>Stories by Rouz Jazayeri on Medium</title>
            <link>https://medium.com/@rouz_85704?source=rss-6a004c556e2d------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 21:04:24 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@rouz_85704/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[Xnor.ai: Transforming AI at the Edge from Hype to Reality]]></title>
            <link>https://medium.com/@rouz_85704/xnor-ai-transforming-ai-at-the-edge-from-hype-to-reality-a607f2c5e638?source=rss-6a004c556e2d------2</link>
            <guid isPermaLink="false">https://medium.com/p/a607f2c5e638</guid>
            <dc:creator><![CDATA[Rouz Jazayeri]]></dc:creator>
            <pubDate>Wed, 12 Feb 2020 18:21:28 GMT</pubDate>
            <atom:updated>2020-02-12T18:36:27.537Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2szo3G_ZLydgziJ4pQIoYg.jpeg" /><figcaption>Xnor.AI Co-Founders: Mohammad Rastegari (left) and Ali Farhadi (right)</figcaption></figure><p>When we started Catapult Ventures, we identified several themes that we wanted to invest behind. Agriculture robotics, high endurance drones for cargo and inspection applications, and new sensors to accelerate the shift to autonomy were just some of the areas that we were (and still are) excited about. But the very first investment thesis we discussed was the need for more efficient AI in order to bring true intelligence to the billions of smartphones, IOT devices, wearables, and endpoints in the world today.</p><h3><strong>Why is AI at the edge so important?</strong></h3><p>Let’s look at an example that we are all familiar with: smart home cameras from the likes of Ring, Nest, Arlo, etc. Until recently, all of these smart home cameras used IR sensors to detect motion and send notifications. While certainly acceptable for the first generation of smart home cameras, dependence on IR triggers a lot of false notifications: tree branches blowing in the wind, car headlights reflecting off of a driveway, or even a spider dropping in front of the camera. This was not a great user experience and also created unnecessary cloud storage costs.</p><p>The first wave of AI-enabled cameras incorporated smart algorithms for things like person detection and package detection. These cameras all sent the images to the cloud in order to properly process the image. This dependence on the cloud created a lot of inefficiencies: poor latency, high cloud costs, poor user experience, and complex privacy, reliability, and security issues. Our cameras certainly got smarter, but they still had a long way to go.</p><p>As we dove into the tech stack and identified the bottlenecks, it became pretty clear that these cameras (and billions of other IOT devices) needed two things: more efficient compute architectures and more efficient inference algorithms to run on these compute architectures. We’ve looked at (and will continue to look at) a lot of companies focusing on efficient AI compute architectures for the edge. But we decided to make an early investment in efficient software knowing that a software-focused company could ramp their business across multiple industries in a short period of time.</p><p>Developing efficient AI at the edge is not a novel concept. A lot of researchers have developed efficient algorithms using binary networks, reduction networks, and other simplifications, but there is an important nuance here: most of those optimization efforts sacrifice performance. The key is to build efficient algorithms that not only minimize compute resources, but also perform with minimal errors. This is where the challenging research problems remain. We quickly determined that these bottlenecks existed in many other verticals like manufacturing, security, automotive, retail, and consumer electronics, which gave us even more conviction to move forward on an investment.</p><h3><strong>XNOR.AI — World Class Efficient AI</strong></h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*KCgQVR96BTm7k8ec934aww.png" /><figcaption>Early Xnor.AI team showcasing their efficient object detection software running on an iPhone.</figcaption></figure><p>Enter Xnor.ai. We first met founders Ali Farhadi and Mohammad Rastegari during one of their visits to the bay area. We were immediately impressed with their technical acumen and it was apparent that they were two of the strongest efficient deep learning researchers we’d ever met.</p><p>We dove into their methodologies for training and inference, and were fascinated that they were able to create high performance, high accuracy inference by careful construction of CNN layers and smart algorithmic optimizations using single bit convolutions. These optimizations can be constructed with only single bit XNOR gates, a fundamental digital logic gate that every sophomore EE student learns about. (I know because I used to teach the digital logic design lab at Purdue.) The performance / watt was so good that they were able to run very advanced deep learning algorithms on low power, efficient Arm processors.</p><p>We were convinced that these were the founders we wanted to back. There was only one problem…despite having roughly half of our fund committed, we hadn’t done a first close yet. We hustled like crazy to finalize the paperwork and closing for our first LPs. Thankfully, Ali and Mohammad gave us some flexibility and a some extra time to get that done.</p><p>Two years later, it’s safe to say that we nailed the thesis. Customers from all types of industries were lining up to do business with Xnor.ai. Some customers wanted Xnor.ai to build customized models and datasets, other customers wanted Xnor.ai to optimize models that were already created. In both cases, Xnor.ai was able to show unbelievable efficiency improvements. I remember in one board meeting, the team showed us how they were able to take a customer’s models for a specific application and improve the efficiency by over 1,000X during a pilot engagement. 1,000X!!!</p><p><a href="https://www.wired.com/story/apples-deal-shows-ai-moving-devices/">Apple recently acquired Xnor</a>.aiand it goes without saying that we are incredibly proud of what Ali, Mohammad, Jon (CEO), and the entire Xnor.ai team have achieved. For Catapult, we still have a long way to go in order to achieve our goals, but this is an early, validating win. We formulated a thesis, we studied the market, we spoke to prospective customers, and then we scoured North America to find the best team. This is the blueprint we follow for a lot of thesis-driven investments. We don’t chase heat, we don’t get hung up on growth/traction/revenue numbers. We invest in founders with deep technical acumen and with a track record for execution in their research and early product development.</p><p>Looking back to our original investment memo, here is an excerpt that we sent to our LPs:</p><blockquote>Finding new technologies that can help accelerate AI inference at the edge is a core thesis for Catapult Ventures. This is where XNOR.AI comes in. Current deep learning algorithms are very compute intensive and require tons of 32-bit matrix multiplication operations. The complexity of these operations requires significant GPU (or CPU) horsepower. XNOR takes these complicated deep learning algorithms and simplifies them down into single-bit (binary) operations, dramatically improving the efficiency of running deep neural nets without sacrificing performance/accuracy. This means that for the first time, AI inference can be run on low power microprocessors or even embedded microcontrollers. This is a very exciting opportunity and it will now allow a whole new level of intelligence in our smartphones, drones, IOT cameras and even across industrial IOT sensors, all places that typically haven’t had the ability to run AI inference locally.</blockquote><p>So, what’s next on the horizon in terms of AI investing for Catapult? We’ll take a couple of months to formulate our thoughts, but it’s pretty clear that Xnor.ai was just scratching the surface. We need more efficient AI tightly coupled with, and maybe even co-designed with, more efficient compute. We need sensors with custom AI for a host of industries that are ready to add new levels of intelligence to their products, equipment, and services. And we need more AI tools to help the next wave of ML engineers accelerate their test-validate-iterate loops.</p><p>Until then, we celebrate and congratulate Ali and Mohammad, the next Chelo-kabob dinner is on us! J</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=a607f2c5e638" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[The Next Wave of Robotics: Autonomous Labor]]></title>
            <link>https://medium.com/@rouz_85704/the-next-wave-of-robotics-autonomous-labor-1ed87426166c?source=rss-6a004c556e2d------2</link>
            <guid isPermaLink="false">https://medium.com/p/1ed87426166c</guid>
            <category><![CDATA[farming]]></category>
            <category><![CDATA[automation]]></category>
            <category><![CDATA[robotics]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Rouz Jazayeri]]></dc:creator>
            <pubDate>Tue, 27 Aug 2019 16:28:47 GMT</pubDate>
            <atom:updated>2019-08-27T16:28:47.577Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*EpDXwNX4mLqLjfZSIlKrGA.jpeg" /></figure><p>As a society, we’ve been infatuated with robots for the past fifty years. From the Jetsons to the Terminator series on the big screen, from robots that vacuum our floors to robots that interact with our children, this obsession has captured the attention and the minds of both entrepreneurs and consumers. Even before the first personal computer was developed, we have dreamed of a future consisting of robots that can co-exist with humans, robots that can help us make informed decisions, and most importantly, robots that can help us complete time consuming and laborious tasks.</p><p>Thanks to billions of dollars of investments into areas like L4/L5 self-driving car technologies and warehouse robotics, the robots are now ready to start working in closer conjunction with humans in a wide variety of fields. Robotic arms have become much more cost competitive, research universities have pushed machine learning and deep learning algorithms to new levels of performance and efficiency, and companies like Nvidia have made AI computing much more pervasive. We are reaching an interesting point where these advancements finally have the potential to augment the labor forces in many traditional industries, and maybe even help solve some significant labor issues that are right around the corner. We call this <em>Autonomous Labor. </em>With the growing number of labor issues surrounding accessibility and affordability, autonomous labor appears to be a fruitful path that is worth exploring.</p><p>Autonomous labor offers the potential for companies to achieve better unit economics by cutting down labor cost and increasing efficiency. However, thanks to Hollywood’s depiction of a world where humans coexist with robots, there has always been a fear surrounding the implementation of automation. The main concerns stem mostly from the threat of “AI software taking over” as well as the concern that robots will eliminate human jobs that are obviously very important to the overall strength of the economy. And in many examples, human coexistence with robots is often not friendly.</p><p>Our take is that automation and robotics should not be utilized to eliminate jobs, but instead, should be used to provide assistance to existing human labor. If implemented correctly, autonomous labor will not only benefit the companies employing them, but could replace traditional back-breaking work with more suitable, labor friendly jobs. We are still in the early days of autonomous labor implementation, but thankfully, there are already many applications across many different industries like farming, long-haul trucking, lawn mowing, warehouse logistics, and even manufacturing. Let’s take a deeper look into one of the areas we are excited about.</p><iframe src="https://cdn.embedly.com/widgets/media.html?src=https%3A%2F%2Fplayer.vimeo.com%2Fvideo%2F352751605%3Fapp_id%3D122963&amp;dntp=1&amp;url=https%3A%2F%2Fvimeo.com%2F352751605%2F3d0daca701&amp;image=https%3A%2F%2Fi.vimeocdn.com%2Fvideo%2F805031202_1280.jpg&amp;key=a19fcc184b9711e1b4764040d3dc5c07&amp;type=text%2Fhtml&amp;schema=vimeo" width="1920" height="1080" frameborder="0" scrolling="no"><a href="https://medium.com/media/1206a5516dfa4d93e78bca07213223c5/href">https://medium.com/media/1206a5516dfa4d93e78bca07213223c5/href</a></iframe><h3>Automation On The Farm</h3><p>In the United States, there are over two million farms that take up over 900 million acres of farmland. Farming accounts for over 47% of the total land in the continental US. And even though there are over a million employees who work on farmlands, some farms still can’t get enough labor to support their harvesting seasons. Rising domestic wages, higher costs and additional complexities associated with foreign labor workers (H2A visa program), and challenges related to finding enough labor workers to harvest on a regular schedule are placing a huge toll on our farmers. This is particularly painful for specialty crop farmers that harvest berries, grapes, lettuce and other fruits and vegetables. These specialty crop farmers pay up to $30,000 per acre per year for labor, or 650 times more than harvesting corn. As a result, many of these farmers are starting to consider contracting the number of acres they plant in the United States and even planting and harvesting more specialty crops in Mexico, which would have a huge impact on specialty crop supply, pricing, and the associated local economies.</p><p>This is an area that we believe is a prime candidate for autonomous labor. Last year, we made a seed investment in <a href="https://advanced.farm/"><strong><em>Advanced Farm</em></strong></a>, whose mission is to build a family of products that can help farmers harvest specialty crops. Their first product, aimed at aiding in the arduous process of harvesting strawberries, is already in deployment on multiple farms in California, the largest strawberry producing state in the country. Their first harvester will use robotic arms and computer vision to identify and pick ripe strawberries more efficiently, and at a bruise rate no worse, than humans. Furthermore, the harvester can run for hours a day and even operate at night (when the berries are firmer and less prone to bruising).</p><p>Autonomous labor will help increase productivity (which is crucial during harvesting season when berries need to be picked every three days) and has the potential to change the labor dynamic on the farm as well as how specialty crops are planted in the fields. Instead of having workers conduct the back-breaking labor of hand picking strawberries in the fields, many jobs could shift towards more of a focus on sorting and packaging, which is a far less physically demanding process and can be accomplished under covering and in the shade. In addition, farmers could plant more crops per acre and increase their output by 15–25% by eliminating a lot of furrows between the planting beds that are needed for carts to move the berries to the trucks. With a robotics solution, strawberries could be picked and transported directly to a central sorting and packing station in preparation for transportation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*p6D72yDHp-85FNUmnsAWew.jpeg" /><figcaption>Field Testing With the Advanced Farm Team at Darrensberries Farm in Nipomo, CA</figcaption></figure><p>Just today, Advanced Farm announced their Series A, with <a href="https://www.ymvsv.com/">Yamaha Motor Ventures</a> as the lead, along with a significant investment from <a href="https://www.kubota.com/">Kubota</a> and a big re-up from <a href="http://www.catapultventures.vc/">Catapult</a>. We are incredibly proud of what this team has accomplished, especially going from concept to commercially viable prototype in an extremely time and capital efficient manner. And we look forward to working closely with <a href="https://www.linkedin.com/in/nolan-paul-538ba12a/">Nolan Paul </a>from YMV, who joins the board of Advanced Farm and just happens to know a thing or two about specialty crops!</p><p>Autonomous labor is an important area where we’ve had strong enough conviction to invest behind some key technology solutions addressing big gaps in some massive industries. In addition to Advanced Farm, we’ve invested in <a href="https://www.starsky.io/"><strong><em>Starsky Robotics</em></strong></a><strong><em> </em></strong>(automated trucking), <a href="http://www.elroyair.com/"><strong><em>Elroy Air </em></strong></a>(autonomous VTOL for cargo delivery), and <a href="https://lefthandrobotics.com/"><strong><em>Left Hand Robotics</em></strong></a> (automated industrial lawn mowing and snow clearing.)</p><p>We continue to be on the lookout for industries that are facing difficult labor dynamics, and specifically, industries that can’t find enough human labor. If an intelligent robotics solution can do repetitive tasks better, do them more consistently, and do them more efficiently, then we’ve found a great candidate for an autonomous labor investment. Our focus will always be to let the robots do what they do best and let the humans do what we do best. And through this dynamic, find the harmony where robots, under the control of humans, and not vice versa, can provide meaningful value to society.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=1ed87426166c" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Catapulting the Next Wave of AI, Automation, Robotics, and IoT Startups]]></title>
            <link>https://medium.com/@rouz_85704/catapulting-the-next-wave-of-ai-automation-robotics-and-iot-startups-c9b977030c8e?source=rss-6a004c556e2d------2</link>
            <guid isPermaLink="false">https://medium.com/p/c9b977030c8e</guid>
            <category><![CDATA[iot]]></category>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[robotics]]></category>
            <category><![CDATA[venture-capital]]></category>
            <category><![CDATA[automation]]></category>
            <dc:creator><![CDATA[Rouz Jazayeri]]></dc:creator>
            <pubDate>Thu, 14 Feb 2019 10:19:13 GMT</pubDate>
            <atom:updated>2019-02-14T10:21:35.396Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*wkH97v-PnLuIFLvZxoqhKw.jpeg" /><figcaption>Darren Liccardo (left) and Rouz Jazayeri (right), Co-Founders of Catapult Ventures, a new seed-stage fund.</figcaption></figure><p>In today’s technology-filled world, buzzwords like <strong>Artificial Intelligence, Machine Learning, </strong>and<strong> Autonomy </strong>have made their way into our everyday lexicon. There are a record number of startups in these areas, with a record amount of capital flowing into these startups. Companies both large and small are attempting to evaluate how these new technologies can help (or potentially disrupt) their businesses.</p><p>The reality is that Artificial Intelligence, Machine Learning, and Autonomy are more than just buzzwords — they are technologies that truly have the potential to alter massive traditional industries like transportation, manufacturing, agriculture, consumer electronics, and life sciences. From self-driving cars and trucks on the roads, to autonomous drones in the air, to robots in the fields, the potential is almost limitless. It’s easy to see why investors are so eager to deploy capital into startups leveraging these trends. We, like other investors, are big believers in these trends.</p><p>While the technology trends that we are targeting are similar, our approach and blueprint are different: 1) Start with a foundational understanding of what is truly state-of-the-art at every level of the technology stack. 2) Identify the current technology gaps that exist in each industry, especially the gaps that, if filled, could lead to large market opportunities. 3) Seek out investments in both horizontal core technologies (applicable to multiple industries) as well as industry-specific vertically-integrated solutions. Both types of investments are needed in order to drive disruptive change.</p><p>So today, we are excited to announce that we’ve launched Catapult Ventures and have completed fundraising for our first $55M fund focused on seed-stage and pre-seed-stage investments in the areas of AI/ML, Automation, Robotics and IoT.</p><p>How did we get here?</p><p>We first met at an industry event about four years ago, when Rouz was working at Kleiner Perkins, doing research on technologies for self-driving cars, and Darren was leading the Autopilot team at Tesla. We stayed in touch, exchanging ideas and references to compelling startups and entrepreneurs in our networks. Later, Rouz and the KP team recruited Darren into DJI, a KP portfolio company, to build out the R&amp;D team in Silicon Valley. As time went on, we collaborated more closely and sourced several compelling seed-stage startups for KP’s Hard-Tech practice. But when some of these very early stage companies didn’t fit the investment profile for KP’s large venture funds, we saw an opportunity.</p><p>We felt that with our complementary skill sets, experiences, networks, and domain expertise, we could offer a unique and differentiated package to entrepreneurs who were just getting ready to launch their companies. In addition to providing early capital, we could help entrepreneurs create their product strategy and technology roadmap, recruit key talent at the right time, and build the most impactful early strategic partnerships. We also wanted to partner with a few corporate investors who could further validate the strategies of our portfolio companies, provide follow-on capital, and become early customers of the companies we back. With Catapult Ventures, we are now doing all of the above.</p><p>Our approach has already resonated strongly with entrepreneurs and we have built a great portfolio aligned with our thesis which includes: <a href="https://www.starsky.io/">Starsky Robotics</a> (autonomous trucking), <a href="https://www.xnor.ai/">XNOR.AI</a> (efficient AI inference at the edge), <a href="http://www.spinlaunch.com/">SpinLaunch</a> (small satellite launching system), Engine ML (accelerating deep learning training), <a href="https://advanced.farm/">Advanced Farm</a> (automated farming), and <a href="http://www.elroyair.com/">Elroy Air </a>(large scale autonomous cargo drones). But this is just the beginning. We plan to build out another 15–20 companies in our portfolio for Catapult’s inaugural fund.</p><p>We are very fortunate to have a fantastic set of Limited Partners, who from the beginning, believed in our vision and strategy. This includes a strong set of corporate investors from the Automotive, Aerospace, and Semiconductor/IoT industries, as well as thoughtful and like-minded institutional and family office investors. Our Limited Partners have enabled us to build out our dream and we are forever grateful.</p><p>As we continue to grow our portfolio, we will never lose sight of our promise to entrepreneurs. We promise to be the most resourceful and most involved early investors in your company. We promise to always have your back and your best interests in mind. And we promise to utilize our network and experiences and to do everything in our power to <em>Catapult</em> your company into the future.</p><p>Looking forward to a great 2019!</p><p>Rouz &amp; Darren</p><p><a href="http://www.catapultventures.vc/">www.catapultventures.vc</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=c9b977030c8e" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>