AI Agents in Production: Why the Platform Matters
summary
Claims that agents need persistent memory, durable execution, real‑time comms and security, supported by concrete primitives: Workers on V8 isolates with near‑zero cold starts, Durable Objects for single‑instance state (demonstrated by a persisted counter), RPC bindings to Workers AI and AI Gateway, browser rendering and dynamic workers for code execution, Facets for sub‑agents, Cloudflare Mesh for on‑prem access, and Artifacts/Artifacts FS for repo per session. Numbers: ~3,000 PoPs, 4.5 million developers (Apr 2026) and 70 models run natively; deployment is one wrangler deploy command.
AI-written summary · about
chapters (10)
- 0:00Cloudflare and agents intro
- 1:22Cloudflare’s platform strengths
- 3:30Workers for scalable agents
- 6:00Durable Objects for state
- 8:23Agents and Cloudflare infra
- 10:52Production agent components
- 14:21Full-stack agent building
- 19:11Deploy agent in 1 command
- 19:33Agent development trends
- 23:26Key takeaways
related talks
15mThe Responsible Path to Hyperproductivity with AI Agents
17mAI TechOps: Data and Infra for AI Foundation to Scale and Manage 1000s of Cohesive Agents
11mCreate AI Agents That Work Independently
05h 17mBuilding AI Applications
26mAI Agents as Your Co-Developer
24mAI Agent Swarms Are Amazing
24mAI Governance for developers: A practical framework for deploying autonomous coding agents safely…- 24mAI Governance for developers: A practical framework for deploying autonomous coding agents safely…