End-to-end RAG system for reliable LLM responses using document retrieval, synthetic data generation, and automated evaluation. Includes vector search, grounded answer generation with citations, hallucination detection, and synthetic Q&A creation to simulate real-world enterprise LLM data and evaluation workflows.

Semantic drift detection for NLP using Sentence Transformers. Detect ML model degradation in production with embedding analysis.