A unified interface to interact with OpenAI, Google Gemini, Anthropic, and other leading providers. Features an intelligent "Auto" mode that dynamically routes queries to the most suitable model based on context and question complexity. Developed with Datastar (Hypermedia/SSE), GO and Tailwind CSS. The platform supports multi-modal analysis (PDF/Image), AI image generation, and semantic session search using Voyage AI embeddings and Turso.

A full-stack data pipeline that transforms a standard Google Drive into a high-performance, searchable knowledge base. It features an automated background worker (Go) that syncs documents, converts them to Markdown, and generates vector embeddings using Voyage AI. The frontend is built with a “No-JS-Framework” architecture using Golang Templ and Datastar, providing real-time, streaming search results via Server-Sent Events (SSE).

Technology video collection with advanced search. Features: Go-powered backend, YouTube API integration for video metadata, dynamic tagging, and vectorization to Pinecone. Implements Gen AI (Open AI) for intelligent query rephrasing and semantic search, all delivered via a Hypermedia-driven UI (Datastar, Tailwind CSS) with YouTube iframe API integration.