Why Faster Simulations Can Change The Way We Build Things | Aaron Wu, CEO & Founder of Khorium AI

Описание: From a student rocket that broke a world record to an AI platform that turns weeks of computer based engineering simulations into hours.

🎙In this episode, I spoke with Aaron Wu, founder & CEO of Khorium AI. His team is building an AI accelerated simulation platform that helps engineers cut work from weeks down to hours by letting AI handle many of the repeated setup steps across different tools.

Aaron grew up in Taiwan, spent ten years playing the trumpet, and won four national championships before shifting into aerospace. At eighteen, he moved to the United States with almost no English and had to learn both the language and the technical basics from free resources like YouTube and Khan Academy, starting at a community college before transferring to the University of Southern California. At USC, he joined the USC Rocket Propulsion Laboratory and led the aerothermal simulation work for the Aftershock 2 mission, where the student team set a world record for college level rocket research with a launch that reached space. That experience showed him how much of innovation depends on being able to test ideas often, without getting blocked by slow and manual simulation workflows.

Building on these lessons, he started Khorium AI. Instead of asking engineers to prepare geometry by hand, create meshes from scratch, enter every boundary condition, and jump between several heavy desktop tools just to run one simulation, Khorium brings the workflow into a single platform and uses AI agents to automate the most complex parts so engineers can spend more time understanding and improving their designs. Aaron also chose to leave a program at Stanford very shortly after starting so he could focus fully on Khorium AI and commit to this problem for the long run.

Learn more: https://www.khorium.ai

Follow Hudson Network for more interviews and startup stories.

💡 Here are some highlights:
→ 🚀 Working on a record setting student rocket showed Aaron how much time teams spend just preparing computer based simulations before they can even test anything in real life. Khorium exists to make that preparation simpler and more structured, so whenever a new idea appears, teams can set up and rerun simulations more easily.
→ ⚡ When weeks of simulation shrink to hours, teams can try many designs instead of betting everything on one choice. Each extra attempt creates more feedback, and that steady flow of feedback leads to better decisions and safer products over time.
→ 🧱 Aaron sees Khorium as a basic brick in the stack for building physical products, not just a niche tool for one small team. The same platform can support aerospace, automotive, energy, and other fields that rely on heavy engineering simulation.
→ 🔍 He believes engineers only trust AI when they can see and check what it is doing. That is why Khorium keeps workflows transparent, with steps and results that can be inspected and verified.
→ 💭 One idea Aaron often returns to is the Top Gun Maverick line “Don’t think, just do.” For him, speeding up simulations is one way to live that out in practice, by letting teams run more experiments, learn from real data, and improve their rockets and products step by step over time.

What I appreciate about Aaron is that instead of accepting the limits in his own work, he chose to build a platform that helps him and other engineers test more designs over time, moving step by step toward safer and more effective products, just like the things we use every day are created through many rounds of testing and iteration.

#KhoriumAI #EngineeringSimulation #Aerospace #CFD #AIToolsForEngineers #DeepTech #StartupStory #ProductivityForEngineers #HudsonNetwork