Serverless single HTML page access to an OpenAI API compatible Local LLM

Talk with EVM Bytecode using webLLM

Designed a secure, AI-based system for real-time fraud detection and behavior analysis. Utilized knowledge distillation, time-series RNNs, and explainability tools (LIME) to ensure efficient and transparent decision-making.

Local Retrieval-Augmented Generation (RAG) system built with FastAPI, integrating vector search, Elasticsearch, and optional web search to power LLM-based intelligent question answering using models like Mistral or GPT-4.

A toolkit of local, privacy-focused AI applications built with Python. Includes a RAG-powered research assistant for PDFs, various chatbots, and visualization scripts.

☕ AI-powered assistant that answers Starbucks-related questions using real reviews, used LLaMA3:Instruct, LangChain, and Ollama , RAG & also Includes real-time Starbucks location lookup, runs fully offline (except for Maps), and styled with a cozy, coffee-themed UI.

AI Think project with ollama local & cloud models

Local ChatGPT using Ollama + Python (Streaming + PDF RAG + Embeddings + FAISS + Chat UI)

Convert Word docs to Markdown privately - 100% offline, no uploads. Perfect for processing sensitive documents with Ollama, LM Studio, GPT4All & other local AI tools. Just double-click the standalone/word-to-markdown.html file to use.

📚 RAGGAE: Retrieval-Augmented Generation architecture for enterprise – Hybrid search, local LLMs (Ollama), NLI compliance, tender/CV/report analysis