1
Production Grade RAG
How to build an integration-enabled RAG chatbot that ingests real-time data from Slack, Google Drive, and Notion.
2
Permissions and Access Control
A deep dive into permissions for RAG - challenges, strategies, and an end-to-end implementation
3
Optimize and Scale Agent tool Calling
Why LLMs degrade under tool overload and the architectural patterns to fix it, with a full implementation across third-party integrations.
4
AI Actions for Workflow Builders
Everything needed to ship a workflow automation interface. Includes: UI, execution engine, integration triggers and actions, and debugging.
Join 2000+ Engineers









CHAPTER 1
Production Grade RAG
Core concepts:
Secure, seamless data access — Enable users to connect their third-party data sources safely (OAuth, permissions, sync).
Permissions-aware retrieval — Enforce each source’s access controls throughout the RAG pipeline.
Fresh, reliable data — Keep indexed content continuously updated so responses stay accurate and production-ready.
CHAPTER 2
Permissions and Access Control for RAG
Core concepts:
Why RAG permissions gets complicated
Walk through different permissions protocols
Implementing a product-ready permissions protocol


CHAPTER 3.1
Optimize and Scale AI Agent Tool Calling
Core concepts:
What it means to be “good” at tool calling
How to improve tool calling
Why agents struggle with tools at scale
How to implement tools at scale
CHAPTER 3.2
Build a Workflow Builder with AI Actions
Core concepts:
Creating a UI for workflow builder
Building workflow execution
Adding integration triggers and actions
Adding self-service debugging







