
Code Story: Insights from Startup Tech Leaders S12 E5: Marc Gyöngyösi, OneTrack
Feb 10, 2026
Marc Gyöngyösi, founder and CEO of OneTrack and ex-autonomous drone builder, turned to computer vision to capture physical‑world data. He talks about pivoting from drones to camera platforms. He explores why warehouses still use legacy systems. He outlines MVP iteration, scaling billions of images, and integrating agentic AI to automate operations.
AI Snips
Chapters
Transcript
Episode notes
From Crashing Drones To Camera Vision
- Marc Gyöngyösi started with autonomous drones in warehouses and even crashed onstage at TechCrunch Disrupt.
- After hardware risks and injuries he pivoted to camera-based computer vision and broader data capture for warehouses.
Ship Many Small MVPs Fast
- Move fast by stacking many small MVPs and iterating daily when possible.
- Build backend flexibility so front-end MVPs can ship without constant rework.
Plan For Rapid Model Change
- Long-term roadmaps break when core tech (LLMs) advances in months or weeks.
- Design modular, repurposable backends so you can swap models and features rapidly.
