AI & MCP Engineering : Build Real AI Features

Everyone is talking about AI. Very few engineers are actually shipping it in production.

I'm an SE-2 at a global product SaaS company where I own an AI chat feature built on LLMs and MCP (Model Context Protocol), currently live and in public preview. I've designed the MCP server architecture, handled real production incidents, and debugged the exact failure patterns that aren't in any tutorial.

This session is for engineers who want to move beyond the hype and understand how AI features actually get built and deployed.

We can cover:

→ What MCP is, how it works, and why it matters for agentic systems

→ How to design and build your first MCP server from scratch

→ RAG pipelines how retrieval, embeddings, and generation fit together

→ Common mistakes in LLM tool integration and how to avoid them

→ Review your existing AI project or idea with honest feedback

This isn't theory from a course. Everything I share comes from real production decisions, real failures, and real architectural trade-offs at scale.

Best for: Backend engineers transitioning into AI, developers starting their first LLM or MCP project, and anyone confused about where agents, tools, and RAG fit together.