- π Currently Building: High-performance Autonomous Multi-Agent architectures and stealth screen-copilots.
- π§ Focus Areas: Multimodal Vision AI, LLM Tool-Use, Model Context Protocol (MCP), and Thread-safe Desktop Automations.
- π οΈ Core Philosophy: Minimal code, zero bloat, high reliability, and low latency.
- π¬ Ask me about: Python, FastAPI, Multimodal AI, NVIDIA NIM, and Agentic Workflows.
| Project | Description | Tech Stack | Status |
|---|---|---|---|
| β‘ ScreenLens-AI | Stealth floating AI desktop copilot that captures screen context and delivers answers in 3 seconds. | Python Vision AI NVIDIA NIM |
π’ Active |
| π€ rishu-payal | Independent Unified AI Agent Platform with modular skill trees and MCP configuration. | Python Multi-Agent MCP |
π’ Active |
| π rishu-antigravity-api | Local API server bridge bringing programmable REST endpoints to the engine. | JavaScript Node.js Express |
π’ Active |
| π² antigravity-telegram-bot | Autonomous AI coding agent bridge running remotely via Telegram interface. | Python Telegram API Gemini |
π’ Active |