RAG and Voice Agent Applications
From Resumes to Conversations: Building Production RAG Systems for Voice-Driven Talent Matching
Discover how to build sophisticated RAG systems that go beyond simple document search to power intelligent voice agents. This session explores real-world challenges and solutions in creating production-grade RAG architectures for complex domains like talent recruitment.
What You'll Learn
Multi-stage RAG architecture: Moving from basic vector similarity to intelligent, domain-aware search with hard constraints, semantic matching, and business logic validation and graph databases
Voice-first data processing: Transforming conversational interviews into structured, searchable knowledge using advanced chunking strategies and multi-modal embeddings
Hybrid search patterns: Combining vector similarity, keyword matching, and structured data filtering for precision at scale
Production challenges: Handling 180k+ profiles with 8 types of embeddings, managing costs, and optimizing for real-time performance
Speaker: Tony Xavier
Resources
- Watched Duration: 21:08
- Watched Duration: 20:27
- Watched Duration: 30:00
- Watched Duration: 27:07
- Watched Duration: 24:07
- Watched Duration: 27:59
- Watched Duration: 29:59
