Proprietary software by InferApps Tech PVT LTD. The source code is private. This page describes the project and what we're building.
One local workbench for a single GPU PC: an agent, unattended pipelines and model management in one place. Every action is recorded with where it came from, so you can see exactly what ran and why.
- Pipelines that survive failure: retries, timeouts, model fallback, crash-safe resume, and human review gates where they matter.
- An agent with three risk modes (ask, gated, auto) and a built-in adversarial reviewer.
- Sandboxed shell commands in containers, with secrets kept in the operating system's credential store.
- Ready-made pipelines: OCR, upscaling, image generation, depth maps, research and knowledge compiling.
- A local dashboard and desktop app, plus a scheduler, MCP tools and reusable skills.
- Works with Ollama, llama.cpp, Lemonade and InvokeAI.
In build.
Questions? Open an issue in this repository. Part of Shanmuga Raja Portfolio.