Becoming an AI engineering company
summary
Claims that companies must adopt an AI engineering paradigm with tight feedback loops and continuous releases; describes an open‑source AI SDK (npm i ai) growing from 150k monthly to ~1.5M weekly downloads and an internal product VZero that nets 100,000 on the wait list in three days. Explains concrete tooling: function calling into React, RAG for proprietary data, evals to catch non‑deterministic outputs, and a 100‑day release cadence proving iterative deployment and rapid user‑driven improvement.
AI-written summary · about
chapters (17)
- 0:00Introduction
- 1:03What Vercel builds
- 2:49AI's impact on software
- 4:33Agentic discovery benefits
- 6:19Vercel's AI framework
- 8:03Demo: AI React components
- 9:24V0 rapid prototyping
- 11:10Academic ML to tools
- 12:36Leveraging AI for business
- 13:36Prompt tuning and RAG
- 15:46Non-deterministic software shift
- 17:10Internal feedback loops
- 18:34Cross-functional AI collaboration
- 19:37Daily feedback integration
- 21:03Shipping fast imperfect
- 21:44Adapting to AI models
- 22:46Community learning outlook
related talks
36mImplementing AI for fun and profit
47mAI, Automation & DevEx: Fueling High-Velocity Engineering
09mAI for React Developers: Opportunities, Learning, and Innovation
28mAI at Work
41m#define AI Engineer
46mBecoming an AI engineering company
27mThe state of AI in software development: Insights across 400+ organizations
39mAI Is Revolutionizing Developer Experience - Are You Ready?