Full-stack application that combines the power of Python and TypeScript. Utilizing Pinecone for vector embedding, Redis for optimized Celery parallel processing, and sql for custom history , users can seamlessly upload, query, and manage PDF documents. The platform incorporates dynamic improvements driven by user feedback, enhancing the querying experience over time.

Features PDF Upload and Query: Users can effortlessly upload and query PDF documents, ensuring a user-friendly experience. It also uses Langfuse for tracing the whole workflow of the following project.

Adaptive Querying: The platform evolves its querying capabilities based on user feedback, continuously improving precision and relevance.

Insightful Metrics: DocumentHub provides valuable metrics, including average scores for Language Model (LLM) performance, database efficiency, and memory utilization.

Tech Stack Languages: Python, TypeScript Databases: Pinecone, Redis, SQLite, SQL Feedback Mechanism: Dynamic adjustments based on user input Metrics: Average scores for LLM models, database performance, and memory usage

Snapshots: