Accelerate AI to prod
summary
Shows how Vercel apps call AWS by configuring OIDC + IAM role setup, setting simple environment variables, then using the AI SDK to import Amazon Bedrock and call generate(text) — one meaningful line of code. Highlights Bedrock features that run inline: guardrails that evaluate responses before return, citations and prompt caching, and token-budgeted reasoning. Announces Agent Core GA as a managed runtime that runs and scales your agent Docker containers, handles identity and memory namespaces, and removes the operational gap to production.
AI-written summary · about
chapters (8)
- 0:00Partnership and background
- 4:47Marketplace and integration
- 6:58The AI developer landscape
- 9:22Integrating AWS Bedrock
- 13:33Advanced Bedrock features
- 19:23Introduction to AgentCore
- 23:10AgentCore architecture
- 27:45Summary and next steps
related talks
25mBuilding Full Stack AI Applications with
28mBuilding Agentic AI applications on AWS
43mBeyond Chatbots: How Amazon Bedrock Agents Drive Enterprise AI
01h 00mSecurely accelerating generative AI innovation
16mEnterprise AI Innovation: Gen AI Experimentation
54mKeynote
46mAccelerate DevOps with generative AI and Amazon CodeCatalyst
32mBuilding a GenAI-Driven Developer Organization