Engineering Production-Scale RAG for Enterprise Java Applications

You'll walk away able to: take your RAG prototype to production, and know when you actually need an agent.

Register NowPresented byNikhilesh Tayal

• ✓Keep retrieval quality, latency and cost under control across millions of documents
• ✓Apply query expansion, cross-encoder re-ranking, metadata filtering and prompt compression
• ✓Design RAG systems that survive failures in production
• ✓Judge how far a well-engineered RAG system can go before you need an AI agent

Your RAG prototype works. Now what?

The engineering challenge begins when the system needs to handle millions of documents, maintain retrieval quality, survive failures, control latency, manage cost, and operate reliably in production.

This session is a practical deep-dive into the engineering techniques that can take RAG beyond a prototype.

We will talk about - query expansion, - cross-encoder re-ranking, - metadata filtering, - prompt compression

And in a world increasingly focused on Agentic AI, we will take a step back and look at an equally important question: how far can we take a well-engineered RAG system before we actually need an AI agent?

Nikhilesh Tayal

Founder & Teacher · AI ML

Nikhilesh is an entrepreneur, teacher, and tech nerd. He is an IIT Kharagpur alumnus and a Google Developer Expert in AI, with over 18,000 followers on LinkedIn.