Chinese IT Association NZ

# Presentation Title:
Beyond just LLMs: Building Physical AI & Robots at Scale

# Abstract:
While Large Language Models (LLMs) dominate today's AI conversations with many people have learned how to ship LLM-powered features quickly, building “embodied” or “physical” AI - robots and intelligent machines that sense, reason, and act reliably at scale in homes or factories presents fundamentally different challenges.

Drawing on real-world product development experience from autonomous home tidy robot to industrial robotic inspection systems, this talk will be a practical look at the constraints shift: latency matters, data is expensive, long-tail edge cases never end, and reliability becomes a product requirement, not a nice-to-have etc.

Key topics include:
- What Physical AI really is: the tech stack makes it work, the control loop, and how it creates complexity that pure software AI products never face.
- Scaling physical AI: the reality gap between demo-ready robot and products that can survive real-world deployment, manufacturing considerations, and quality control when every unit must perform consistently.
- The product manager's perspective: unique challenges of hardware-software synthesis, common failure modes, and trade-offs between capability, reliability, safety and cost.

Whether you're curious about robotics, exploring career paths beyond pure software, or interested in how AI is reshaping the physical world, this session offers practical insights from the frontlines of building AI that moves.