Technion Trailblazers: Jonathan Spitz on Transforming Industry with AI
The magnitude of artificial intelligence’s potential can cloud our imagination when trying to understand its specific real-world impact. Fortunately, the Technion – Israel Institute of Technology and its expansive network of alumni around the world are working each day to ensure that its potential is honed for good and advance true transformation that we can see with our own eyes.
One such alum is Jonathan Spitz, founder of AI start-up GaussML. He leverages physical AI—the new frontier of AI innovation that will bring about a future where robots are our helpful friends, rather than the villains of science fiction. I recently spoke with Jonathan about what inspired him to bring much-needed new technology to an overlooked industry and how his years at the Technion shaped his future.
Michael: Knowing you’ve been acquainted with the American Technion Society (ATS) in the past, it’s great to reconnect and chat in more depth! To reflect on when you started at the Technion many years ago, can you tell me more about what inspired you to pursue your studies in Haifa?
Jonathan: Yes, it’s great to have the opportunity to share my story! I remember fondly my internship in Palo Alto while I was a student at the Technion and my first introduction to ATS while I was there.
I’ve spent much of my life bouncing around the globe. I’m originally from Argentina, and I moved to Israel when I was just 17 years old. Attending the Technion had been a dream of mine for years, and it was a significant factor in my decision to start a new life in Israel. Its reputation for academic excellence and access to industry connections echoes around the world.
Even beyond the Technion itself, I knew that Israel offered countless opportunities for my future. As Start-Up Nation and a global tech hub, Israel is the place to be if you want to make a real impact on how people live in this technological age.
Michael: The Technion’s history as a driver for innovation has shaped life in Israel itself for more than a century and has impacted technological advancements worldwide. With research and innovation moving so quickly, did you know which area you wanted to pursue?
Jonathan: I was ecstatic to have earned top grades in high school which opened up several options to choose from when selecting a Faculty I wanted to study with. I had a plan to pursue robotics, so I kicked off my education in the Faculty of Mechanical Engineering. I feel very lucky to have found my passion right away.
Wanting to learn everything I possibly could about robotics and how to train them to do exactly what I wanted, I remained at the Technion through my Ph.D in Mechatronics, Robotics, and Automation Engineering. In my thesis, I created a robot that could walk across different conditions of terrain without requiring a ton of feedback to refine its ability.
Michael: As one of Technion’s AI trailblazers working to shape this new reality worldwide, do you feel like the Technion gave you the background needed to embrace this rapid progress today?
Jonathan: AI was in its very early stages and had limited applications during my studies at the Technion. At the time, it was simply called machine learning. A huge part of my education centered on machine learning and involved writing algorithms so that robots knew what to do and how to react to new environments.
Say a robot is walking across a field and encounters a tree. My algorithm must be designed so that the robot not only recognizes the tree but knows how to avoid it as well. That image processing requires simple machine learning. If you think about it, simple robots like that serve as a micro-example of how large language models (LLMs) like ChatGPT work today. You feed the chatbot information that it must process to understand, and then it creates a solution based on its own analysis. Of course, the actual mechanics of AI models are much more complex than that, but my educational foundation in robotics provided a great launchpad for what I do now.
Michael: From my conversations with other alumni, I know that launching a start-up can be incredibly daunting. Sure, Start-Up Nation lends itself to inspiration around every corner, but I imagine it felt like a risk.
Jonathan: As you well know, it’s almost impossible to live in Israel without feeling the pressure to start your own company. Israel and entrepreneurship are essentially synonymous. I dipped my toes into entrepreneurship for the first time while at the Technion participating in the annual BizTech competition. I was pursuing my master’s degree at the time and built a small team of 3D printed robots which were featured in TechCrunch. We thought they may have applications in education, but long story short, we were unsuccessful in getting the idea off the ground. That said, the experience left a lingering willingness and motivation to try again.
To your point, launching a start-up is terrifying and exciting all the same. Before biting off more than I could chew, I decided I required some more experience in the professional world before trying to build an empire of my own. After completing a post-doc program at Inria, the French national institute for research in AI, I secured a role at Bosch in Germany as a research scientist. There, I helped develop machine learning algorithms for industrial optimization.
With two years of valuable experience on the corporate side under my belt, I felt much more confident to dive back into entrepreneurship. In 2020, I left Bosch and launched my startup, GaussML.
Michael: Like many Technion alumni, your journey started with several twists and turns. That depth of experience really highlights the range of impact that Technion graduates can have in the real world. What factors informed your decision to focus on manufacturing with your start-up?
Jonathan: To give some context on what inspired my pivot, when I started GaussML in 2020, the tech world was already consumed by the excitement and possibilities of LLMs that were then introduced to the public a few years ago. What was starting to pick up steam but hadn’t yet broken through to the mainstream was physical AI. LLMs are trained on data that exist on the Internet, but physical AI models are trained and adapted based on information it gathers in the real world. Consider humanoid robots we occasionally see in the news that look futuristic yet struggle with everyday tasks like loading a dishwasher. While still far from having a meaningful impact on our day-to-day lives, they serve as great examples of what physical AI might look like in a more consumer-facing application several years from now.
However, there are industries that don’t need to wait to see the transformative impact of physical AI. That’s where GaussML comes in. Every component of my background up to 2020 pointed at the next frontier of manufacturing. With my education in machine learning and robotics, professional experience applying my skills to manufacturing at Bosch, and training in Start-Up Nation, launching an AI-powered manufacturing company made perfect sense.
Michael: That’s impressive to take on building a company based on technology that is still in its very early stages. How exactly does GaussML hope to transform this industry? I could see it having an incredibly transformative downstream impact on the economy.
Jonathan: Traditionally in manufacturing, the machines on the production line require extremely precise settings to run correctly and efficiently. Those instructions occasionally need fine tuning, but there’s a global lack of specialized talent that has the expertise required to properly operate those machines. Companies spend millions of dollars on their manufacturing machinery, and don’t even use them to their fullest potential.
With GaussML software, companies can use our technology to unlock the correct parameters and instructions that allow their machines to operate at full capacity. We’ve worked with household names like Coca-Cola to optimize their plastic injection molding and reduced their production time by 10%. We also worked on a project with a global leader in electric vehicles, and we’re looking forward to launching a partnership with one of the largest supermarket chains in Germany later this year with our technology on their product lines.
You’re correct that if implemented widely, physical AI can make manufacturers more profitable and improve the supply chain. If companies can operate more efficiently, margins improve, and benefits trickle down to every corner of the economy.
Michael: That’s a great way to put into perspective the real-world impact of AI on our society. It’s easy to talk about AI and its potential in sweeping terms, but knowing what that means in real life is vital for us to know how and when to apply it properly. Can you discuss more about how you see AI transforming your industry?
Jonathan: AI as we know it today has huge, untapped potential. We can only imagine how it will improve and change, but I can say with almost one hundred percent certainty that AI will transform almost every industry. It’s just a question of how long that will take.
For manufacturing, most of the industry is lagging. In fact, I’m sometimes nervous that GaussML is too early for the market. Right now, in most manufacturing contexts, companies are gathering data from their machines and uploading it to a dashboard, trying to make connections on their own. We provide an alternative: companies don’t need to do that work alone. Not only can we save the time of rummaging through data in hopes of finding something useful, but we also actually help transform the machinery to run more efficiently.
I’m certain we’ll see more competition down the line, and I’m excited that we can be a first mover in applying physical AI to a new area. Physical AI has a bottleneck in available data right now, which is why it hasn’t taken off in the same way as LLMs just yet. However, I know it will have a significant impact soon since innovation is progressing every day.
Michael: It sounds like GaussML is long overdue, and I couldn’t be more delighted to see a Technion alum bring a visionary idea to life. I’ll be cheering you on and cannot wait to see how it continues to grow. Looking forward, and knowing you’re quite the globe trotter, do you still feel connected to the Technion today?
Jonathan: The Technion was my life for over a decade, and now several years after I completed my studies, I still use skills I learned there every day. I always say that my startup and daughter were born just a week apart, and it’s a reminder that life moves very quickly. Recently, the Technion Alumni Association has pushed to created communities of alumni around the world, and it’s been so rewarding to connect with my fellow Technion peers in Germany and celebrate new parts of my life with people who share a similar past. Remembering the Technion foundation that unlocked the opportunities that have defined my life so far consistently reminds me that I will always have a home and community in Haifa.
This Q&A was made possible by collaboration with the Technion Alumni Association.
Michael is the out-going CEO of the American Technion Society, and this blog reflects his own personal opinions, and not the ATS or Technion.
