Agentic AI, LLM engineering, and the models behind modern automation — multi-agent systems, LLM evaluation and comparisons, RAG, fine-tuning, AI infrastructure, security, and production AI engineering.
From the blog
Master FastAPI's Depends system to inject LLM clients, database sessions, and agent configurations into your AI agent endpoints. Covers scoped dependencies, sub-dependencies, and testing with overrides.
Build a production middleware stack for AI agent APIs in FastAPI. Covers structured request logging, Bearer token authentication, sliding window rate limiting, and CORS configuration for agent frontends.
Write comprehensive tests for AI agent APIs using pytest and httpx. Covers TestClient usage, async test patterns, fixture design for database and LLM mocking, and strategies for testing streaming endpoints.
Learn how to implement feature flag patterns for AI agents including percentage-based rollouts, user targeting, and kill switches. A practical guide to safely shipping new agent behaviors to production.
Build secure file system tools that let AI agents read, write, list, and manage files within a sandboxed directory. Covers path validation, sandboxing strategies, permissions, and safety considerations for production deployment.
Handle file uploads in FastAPI for AI agent document processing and image analysis. Learn type validation, size limits, chunked uploads for large files, and async processing pipelines for uploaded content.
Design and implement fine-grained permission systems for AI agent tools using RBAC, ABAC, and policy evaluation. Includes FastAPI examples for dynamic, context-aware access control.
Learn a practical decision framework for choosing between prompt engineering, retrieval-augmented generation, and fine-tuning based on cost, data requirements, latency, and use case complexity.
A hands-on tutorial for fine-tuning open-source LLMs using Hugging Face Transformers, PEFT, and TRL libraries, covering setup, training configuration, evaluation, and pushing to the Hugging Face Hub.
