LLM Analytics Dashboard

Real-time LLM performance dashboard with RAG capabilities and document analysis

LLMRAGAIAnalyticsDashboardWeb App

About This Project

An interactive dashboard showcasing real-time analysis of LLM (Large Language Model) performance metrics including token usage, latency, speed, and cost tracking. The platform features advanced RAG (Retrieval-Augmented Generation) capabilities, allowing users to upload documents (PDF, DOCX, TXT) which are automatically chunked and embedded using Gemini AI. The dashboard provides semantic search functionality to find relevant document chunks for queries, displays used sources with expandable sections, and supports multiple Gemini models (2.5 Flash and Pro). Built with Next.js, it features streaming responses, markdown rendering, and a modern, responsive UI.

Key Features

Real-time LLM metrics tracking (tokens, latency, speed, cost)

Document upload and automatic embedding generation

Semantic search with Gemini embeddings

Interactive chunk viewer with expandable sources

Support for multiple Gemini models (Flash & Pro)

Streaming responses with markdown rendering

Drag-and-drop file upload interface

Technologies Used

Next.js

React

TypeScript

Tailwind CSS

Gemini API

pdfjs-dist

mammoth

Framer Motion

Back to Home