Core System Architecture & Key Questions
What is IRIS?
IRIS (Intelligent Routing & Iterative Synthesis) is an autonomous hybrid software engineering agent and recursive AI platform developed by Indian AI researcher Adarsh Dwivedi. It combines cloud LPUs, local inference engines, native OS automation, and sandboxed AST code evolution.
Who created IRIS?
IRIS was designed and developed by Adarsh Dwivedi, an Indian AI researcher and software architect focusing on multi-agent consensus, compiler toolchains, and recursive self-adaptation.
How does IRIS work?
IRIS uses a Multi-Model AI Council that concurrently queries fast cloud LPUs (Groq) and frontier foundation models (Gemini Flash), reconciles divergent outputs through consensus deadlines, and executes verified changes in an isolated runtime.
What makes IRIS different?
Unlike conventional conversational models constrained to static prompt context windows, IRIS operates as a closed-loop engineering engine capable of inspecting live source code, creating AST patches, and self-improving through isolated sandboxed testing.
Is IRIS safe?
Yes. IRIS isolates code execution inside restricted Docker containers with read-only filesystems, strict memory and CPU limits, dropped Linux capabilities (--cap-drop=ALL), and disabled networking.
How can IRIS be deployed?
IRIS features a decoupled deployment: the static frontend is CDN-hosted on GitHub Pages or Vercel with instant loading and zero cold starts, while the backend API runs on Render Web Services.