From Self-Driving Monorepo to Self-Driving Cars
AI coding agents promise massive productivity gains, but realizing that promise at scale requires more than just tools. In this talk, I’ll share how we approach AI adoption at Zoox, including: - Designing a monorepo-friendly ecosystem of agents, tools, and workflows - Driving adoption through enablement, hackathons, and internal platforms - Defining and tracking meaningful productivity metrics beyond hype - Managing token spend and aligning it with business outcomes - Structuring Skills, CLIs, MCPs, and Plugins to scale across teams The goal is simple: turn AI from an experiment into a reliable, measurable, and scalable engineering capability.
About the Robotics & World Models Track
Robotics & World Models sessions at AI Engineer World's Fair 2026 in San Francisco.
When
Wednesday, July 1, 2026
3:20 PM - 3:40 PM·20m
Where
Track 2 · Room 2006
Capacity: 250 attendees
Sign in to add this talk to your schedule.
Speaker
Amit Navindgi
Zoox
@amitnavindgi
Amit Navindgi is a Senior Staff Software Engineer and AI lead at Zoox, where he founded and leads Zoox Intelligence, a company-wide initiative applying Large Language Models across engineering, operations, customer support, autonomy, and employee productivity. His work combines platform engineering with organizational AI adoption. He architects internal AI platforms, agents, and developer productivity workflows, while also leading AI tool evaluation, rollout strategy, enablement, spend management, and productivity measurement across Zoox. Amit also runs the Zoox Hackathon and The Assembly, a cross-functional forum for knowledge sharing and innovation. Earlier in his career, he built web applications and distributed systems at Veritas Technologies and worked on Natural Language Processing at Xerox Research Centre Europe.