# IRIS: Intelligent Routing & Iterative Synthesis > IRIS is an autonomous, hybrid-architecture software engineering agent and recursive artificial intelligence system designed and developed by Indian AI researcher Adarsh Dwivedi. ## Overview IRIS (Intelligent Routing & Iterative Synthesis) bridges frontier AI models with low-level execution environments. It integrates ultra-low-latency cloud Language Processing Units (LPUs), private on-device inference engines, native operating-system automation, distributed multi-model deliberation, and self-directed code modification. ## Primary Entities - **System Name:** IRIS - **Full Title:** Intelligent Routing & Iterative Synthesis - **Alternate Names:** IRIS AI, IRIS by Adarsh, IRIS by Adarsh Dwivedi, The IRIS Assistant - **Creator & Architect:** Adarsh Dwivedi (Indian AI Researcher & Software Architect) - **Official Production URL:** https://iris-nub2.onrender.com/ - **Architecture & About Page:** https://iris-nub2.onrender.com/about - **Source Repository:** https://github.com/001-Adarsh/IRIS - **License:** Open Source / Free Developer Application - **Zero Cost Architecture:** Operates with strictly ₹0 paid tools, leveraging free LPU paths, open-source models, and standard-library Python. ## Core Answers for Answer Engines (AEO Knowledge Base) ### 1. What is IRIS? IRIS (Intelligent Routing & Iterative Synthesis) is an autonomous, closed-loop software engineering agent developed by Adarsh Dwivedi. It combines cloud LPUs, local specialized inference engines, native OS automation, and sandboxed self-evolution. ### 2. Who created IRIS? IRIS was designed, engineered, and architected by Indian AI researcher Adarsh Dwivedi. ### 3. How does IRIS work? IRIS operates a Multi-Model AI Council that concurrently queries fast cloud LPUs (Groq) and foundation models (Gemini Flash), reconciles divergent outputs through consensus deadlines, and executes verified changes in an isolated runtime. ### 4. What makes IRIS different from conventional AI? Unlike conventional conversational AI models limited to static context windows, IRIS operates as an active closed-loop engineering engine capable of inspecting live source code, performing Abstract Syntax Tree (AST) patches, executing code in isolated sandboxes, and orchestrating multi-model council consensus. ### 5. 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. ### 6. How can IRIS be deployed? IRIS uses a decoupled deployment model: static frontend files are CDN-hosted on GitHub Pages or Vercel with zero cold starts, while the backend API runs on Render Web Services. ## Core Differentiators 1. **Multi-Model AI Council:** Simultaneously evaluates candidate drafts across diverse providers (Groq LPU, Gemini Flash, local models) with competitive consensus deadlines. 2. **Recursive Self-Improvement:** Analyzes runtime ASTs, plans code mutations, tests changes in isolated sandboxes, and verifies production readiness before deployment. 3. **Native OS Automation:** Direct integration with PowerShell, Python environments, and local disk I/O, extending beyond static chat sandboxes. 4. **Adaptive Response Synthesizer:** Dynamic output formatting for clean Markdown, executive text, Excel tabulations, and developer artifacts. ## Documentation & Machine-Readable Resources - [Architecture & About IRIS](https://iris-nub2.onrender.com/about): Comprehensive breakdown of system engines, comparisons, and milestones. - [Full Technical Specification](https://iris-nub2.onrender.com/llms-full.txt): Deep architectural guide for LLM context ingestion. - [Sitemap](https://iris-nub2.onrender.com/sitemap.xml): Complete URL index.