Sleek Streamlit chat app for Google Gemini (Files API). Dark, gradient UI with model picker, usage dialog, file/image/audio/PDF attach & preview, chat history, image persistence, robust error handling, and token usage tracking. Supports streaming replies and modular backend via google-genai.

RAG-based chat application using Amazon Bedrock, Aurora Serverless, Terraform, and Streamlit.

An end-to-end Generative AI RAG project built on AWS Bedrock, Aurora Serverless PostgreSQL, Amazon S3, and Terraform, with a Streamlit chat frontend.

AWS-native document Q&A. Documents live in S3; a Bedrock Knowledge Base (Aurora pgvector) backs retrieval; a Bedrock Agent runs a multi-step loop (retrieve, re-query, call tools) instead of one-shot RAG. A Streamlit UI on ECS sits behind an ALB; chat goes to a Lambda that invokes the agent.