context-aware-ai-assistance

Context-Aware AI Assistant is an on-screen AI assistant that understands the context a user is currently looking at. It allows users to select text directly inside an application, click Ask AI, and receive a contextual response without leaving their current workflow. The application provides context to the AI instead of requiring the user to explain that context manually.

The project uses Retrieval-Augmented Generation (RAG) with MongoDB Atlas Vector Search and Gemini integration. It supports application-specific knowledge, file-based knowledge ingestion, configurable system prompts, context-grounded responses, and follow-up questions. The RAG pipeline includes document loading, text splitting, embedding generation, semantic retrieval, and grounded response generation.

Setup involves configuring application.properties with Gemini and MongoDB Atlas credentials, updating a system prompt file, and starting the Spring Boot application. Users upload application documentation or source files via POST /upload, then send queries via POST /query?q=YOUR_QUESTION. The assistant retrieves relevant context from the vector store and generates a response based on that context.