Building Deep Tech with Ilir Aliu

Steve Xie is the founder and CEO of Lightwheel AI, building the simulation and synthetic data layer powering the next generation of embodied AI and humanoid robotics.In this episode, Steve shares a rare founder journey that starts far from robotics. From studying physics at Peking University, struggling to stand out, and rebuilding confidence through sheer consistency, to a PhD at Columbia and an early failed startup built out of love for his dog. A detour that taught him the cost of building without a business model.We talk about his path through Cruise, NVIDIA, and NIO, where he led large-scale simulation efforts for autonomous driving. Steve explains how those years shaped his conviction that simulation, data quality, and evaluation are the real bottlenecks in physical AI.He then breaks down how Lightwheel AI came to life. Why sim-ready assets matter more than solvers. How synthetic data actually closes the sim-to-real gap. And why robotics teams hit a ceiling without proper evaluation and scaling infrastructure.A deep conversation about resilience, delayed gratification, and why the hardest part of building is often unlearning what made you successful before.

I talk with Animesh Garg,Assistant Professor at Georgia Tech and one of the leading voices in robot learning today:We talk about growing up in India, building his first autonomous vehicle on a $280 budget after being rejected from a CMU program, and how that failure pushed him toward Berkeley, Stanford, and eventually NVIDIA Research. Animesh shares why he avoided computer science early on, what drew him to mechatronics, and how curiosity rather than planning shaped his entire career.ORBIT and Isaac Lab, why simulation is now the backbone of robot learning, and how world models, reinforcement learning, and foundation models are lowering the barrier for people outside robotics to build real systems. Animesh explains why he believes the most important robotics breakthroughs will come from people who are still in high school today.A deeply personal conversation about grit, risk, redefining success, and why chasing interesting problems beats chasing money.

In this episode, I talk with Prof. Dr. Marco Huber, Professor for Cognitive Production Systems at the University of Stuttgart and Scientific Director for AI at Fraunhofer IPA.Marco shares his journey from a middle-class upbringing with no academic role models to becoming a leading figure in applied AI for manufacturing. We talk about discovering computer science through a single physics teacher, why he almost went to vocational school, and how a mix of personal drive and mentors shaped his path.We spend some time on his years between academia and industry, what he learned working in high-pressure startups, and why real innovation happens when theory meets factory floors. Marco explains how Germany still leads in fundamental AI research but is at risk of losing the race when it comes to turning research into scalable industrial products.A conversation about explainable AI, robotics in production, and why Europe has only a small window left to turn Physical AI into a competitive advantage.

Stefan Dörr-Laukien, CEO of NODE Robotics, shares his journey from mechanical engineering to robotics entrepreneurship. He discusses the critical transition from research to real-world applications, emphasizing market needs over perfect technology. Stefan highlights challenges in scaling mobile robots and the importance of modular, hardware-agnostic software. The episode delves into the risks of being too technology-focused, decision-making in uncertainty, and customer trust in innovative solutions. It's an insightful look into navigating the robotics industry's complexities.

I talked with Stephan van den Brink, founder and CEO of MANUS™, the company behind some of the most advanced data gloves used in robotics:Not only robotics: teleoperation, motion capture, and embodied AI. Manus started as a small student project and grew into a deep tech company trusted across the robotics world.We talk about Stephan's path from studying law and economics to discovering he was meant to build things, not file documents. He explains how the first Manus glove was built in evenings and weekends, how an early Kickstarter failure opened better doors, and how an accelerator program became the turning point for the company.Stephan shares how Manus shifted from VR gaming to B2B simulation, then to motion capture, and now to robotics. He explains why EMF tracking became their core technology and why precise hand data is suddenly in huge demand as humanoids and AI driven robots take off.We also talk about building a company for ten years, staying alive through hype cycles, making hard calls, and focusing on what real customers need.

In this episode, I talk with Harald Schäfer, CTO at comma.ai, where he is leading one of the most interesting autonomy efforts in the world:They work on end to end driving and generative world models is changing how small teams can compete with billion dollar labs.We talk about his path from electrical engineering in Belgium and Santa Barbara to joining comma as one of the earliest engineers. Harald explains how he helped turn a hacker project into a focused engineering team that ships reliable autonomy to thousands of real users.He walks me through comma’s move to a single neural network that controls the car from video input, why deleting code is often more powerful than adding more, and how his team uses world models to train on billions of synthetic miles that never existed on real roads. Harald also shares what it is like to build inside a company with no CEO, why simplicity beats complexity in autonomy systems, and how the new comma 4 and Body 2 signal a move beyond cars into general robotics.If you work in robotics, autonomy, or AI systems, this conversation is packed with lessons about engineering clarity, avoiding brittle stacks, and shipping real products with small teams.

Albane Dersy turned down Goldman Sachs to build Inbolt, a robotics company now deployed in factories across the world. Her story is a masterclass in execution:In this episode, we talk about how Albane grew up in Paris, pushed her way through the French prep school system, and found her path into entrepreneurship after a semester at Wharton opened her eyes to what was possible. She explains how she met her two co-founders during the X EC program, and how the first version of Inbolt had nothing to do with robots. They started with a real-time guidance tool for workers and later pivoted to industrial robots after spending months on factory floors and seeing where customers really needed help.Albane walks through what it takes to sell and deploy automation inside global companies. She talks about why founders need to be on site all the time, and why selling early matters more than waiting for perfect reliability. She explains why deployment is everything in manufacturing and how Inbolt built a system that retrofit existing robots, reduced downtime, and proved value in a few weeks instead of years.We also talk about ambition, hard work, and the pressure she faced breaking into industries that are not always welcoming to young founders. Albane shares her early years in boxing gyms, her drive to be taken seriously, and the mindset that helped her operate and grow a company that now works with some of the biggest manufacturers in the world.If you want a clear look at how real robotics gets deployed at scale, and what it takes to build a company inside the most demanding industry in the world, this is an episode you should hear.

Dr. Hendrik Susemihl, CEO and Co founder of GoodBytz, shows you how fully automated kitchens can solve the labor crisis in food service and still serve better, fresher food at scale.We talk about his path from taking apart PCs as a teenager, to building large automation systems at Fraunhofer, to becoming CTO at NEURA Robotics. Hendrik explains why he walked away from a safe leadership role after his father’s heart attacks, how going plant based changed how he sees food, and why he became obsessed with the question: if I can cook healthy meals quickly at home, why is it so hard to get that quality in hospitals, canteens, and on the road.Hendrik breaks down how GoodBytz works in practice: a compact robotic kitchen that cooks up to 150 meals per hour, runs 24/7, and delivers consistent quality in places like university hospitals and motorway sites. We get into what they learned from running their own Lieferando brand, why he mostly ignores CVs and hires for people who build things for fun, and how a small Hamburg startup ended up signing a landmark contract with the US Army to feed soldiers in South Korea.If you care about robotics with real deployment, food at scale, or building a deep tech company that actually ships, this episode will be very useful for you.

In this episode, I talk with Jon Miller Schwartz, co-founder and CEO of Ultra, about how to actually get robots deployed in warehouses:We walk through Jon’s journey from tearing apart electronics on a tiny New York City workbench to Harvey Mudd, early YC startups in 3D printing, and building one of the first highly automated factories at Voodoo Manufacturing. Jon explains why those painful years with “last generation” robots convinced him to start Ultra and focus on one thing first e commerce order packing as a beachhead for real industrial deployment.He breaks down how Ultra’s robots drop into existing pack stations, learn from examples instead of brittle scripts, and why he believes in multi purpose robots before truly general purpose systems. We talk about force sensitive dexterity, what most people get wrong about warehouse automation, and how a small team in Brooklyn already has robots running live for customers. If you care about turning AI and robotics into shipped systems instead of slideware, this one is for you.

Axel Peytavin, co-founder and CEO of Innate, specializes in making personal robotics accessible to everyone. He shares his journey from France to Stanford, revealing how early coding shaped his path. Axel discusses Innate’s new $2K teachable robot, which can learn skills rapidly, highlighting its real-world applications like security patrols and chess playing. He emphasizes the importance of speed in innovation and how personal robots could revolutionize technology in a way similar to personal computers, advocating for an open, collaborative approach in the field.