Every MCP starter on GitHub is a hello-world. I needed production features. I spent the last few...
Full walkthrough: architecture, CDK deployment, real cost breakdown, and performance numbers from production. Open source.
The fun part is building the AI. The hard part is not going broke - lessons from production cost tracking, security, and scaling
How I built a TypeScript SDK that makes AI platform integration feel effortless - complete with streaming, types, and error handling
Why I moved from Lambda functions to ECS Fargate for AI agent orchestration - and how the ReAct pattern changed everything
Most RAG tutorials use 20 lines of LangChain and call it done. Here's how to build RAG that works on real documents in production.
How we unified 7 AI Lambdas into one API that works with OpenAI, Claude, Bedrock, or any provider with a simple config change.
Lambdas seemed perfect for AI workloads. Then we hit the 29-second timeout, cold starts, and cost spikes. Here's what we learned.
Managing dozens of AI Lambdas taught me that existing solutions aren't built for real-world production. Here's why I built my own.
I've built the same serverless API setup at least five times now. I work with AWS and TypeScript...
I've written the same NestJS CRUD code at least 50 times. Controller with five endpoints. Service...
Every project I've worked on eventually needs to import or export Excel files. Customer data,...
Every time I start a new microservices project, the same thing happens. I spend the first three...