build-on-aws/langchain-embeddings

Supports text, image, and video content indexing through multimodal embeddings (Nova models), with production-ready architectures using Aurora PostgreSQL with pgvector for similarity search and Lambda/ECS for serverless or containerized document processing. Includes progressive learning path from Jupyter notebooks covering semantic search fundamentals through advanced agentic RAG systems, integrating AWS services like Bedrock, Transcribe, and Step Functions for end-to-end workflows.