Tool or Die x Maneva

Tool or Die x Rae Jeong: From Welding to DeepMind to Revolutionizing Manufacturing with AI

What does a childhood spent welding in small-town Alberta have to do with cutting-edge industrial AI?

In a wide-ranging episode of Tool or Die, Maneva co-founder and CEO Rae Jeong shares how his path—from working in a mechanic’s shop to AI research at Google DeepMind—led to building Maneva, a company that’s redefining how factories operate using video-based AI agents that live on the factory floor.

Listen to the episode‍

From Blue-Collar Beginnings to Reinforcement Learning at the Edge

Maneva was built around a simple but powerful idea: if a human can look at something on the factory floor and make a judgment, then so can an AI — if it sees video, not just images.

Unlike traditional systems that require ideal lighting, expensive hardware, or major infrastructure overhauls, Maneva’s video-to-action AI uses reinforcement learning to adapt to real-world environments, working seamlessly with existing equipment.

From jammed conveyors and defective packaging to safety lapses and root cause analysis, Maneva’s AI agents don’t just observe. They take action.

What You'll Hear in the Episode

• Rae’s path from South Korea to Leduc, Alberta
• How working with his hands shaped his view of AI in industry
• Why DeepMind’s cutting-edge research wasn’t enough — and what drove him to leave
• The core of Maneva’s product: video AI agents that act in the real world
• How reinforcement learning enables plug-and-play deployment at the edge
• A deep dive into Kaizen 2.0: factory-wide root cause analysis powered by AI
• Real-world examples: catching jammed trays, broken prep processes, and more
• The philosophy behind building mission-critical AI that respects the operator and doesn't just monitor — it protects

Not Just Vision. Video.

Most industrial AI claims to use “vision.” Maneva chose the word “video” for a reason. The AI needs to understand time, not just freeze frames. Whether it's watching the sequence of a safety protocol or tracking product flow, video allows for nuance, context, and intervention.

“If you can see it in a video, our AI can learn it, understand it, and act on it.”
— Rae Jeong, CEO of Maneva

More Than Just Quality Control

While Maneva’s AI excels at defect detection, that's just the beginning. The system also helps with:

• Downtime prevention — spotting jams, product misfeeds, or early signs of equipment failure
• Safety monitoring — alerting operators in real time, not punishing them afterward
• Root cause analysis — giving plant leaders visibility into bottlenecks across multiple stations
• Process optimization — identifying slowdowns or compliance gaps before they impact output

All of this happens without requiring factories to change their equipment, their layout, or their processes.

Why This Matters

Manufacturers today are navigating labor shortages, aging equipment, and rising complexity. Most can’t afford to rebuild their plants from scratch. Maneva helps them layer intelligence onto what they already have — turning their lines into adaptable, responsive systems without disrupting operations.

It's not about replacing workers. It's about helping them do more, safely and efficiently.

And if you're in manufacturing, AI, or just love a good founder story,
listen to the full conversation with Rae on Tool or Die