IRIS Autonomous Engine
A hybrid software engineering system featuring multi-model council deliberation (cloud LPUs and local models), dynamic AST code self-mutation, and isolated Docker sandboxing with zero host compromise.
Building autonomous software architectures, scalable operational engines, and high-precision systems.
I am a builder and systems architect driven by a simple obsession: turning complex, high-friction processes into elegant, self-sustaining systems.
With an engineering foundation rooted in mechanical first principles, I look at software not just as syntax, but as a dynamic machine of throughput, feedback loops, and deterministic precision. Whether I am architecting autonomous agent loops or governing large-scale enterprise data ecosystems, my core instinct is to build software that creates real leverage.
I created IRIS (Intelligent Routing & Iterative Synthesis) to bridge the gap between conversational AI and real software engineering. Rather than treating artificial intelligence as a passive text generator, IRIS is engineered to inspect live codebases, mutate abstract syntax trees (ASTs), validate patches inside isolated execution sandboxes, and orchestrate multi-model council deliberation.
When I am not designing software architectures, I explore autonomous systems, high-performance data modeling, and tools that empower creators to operate with unprecedented speed and clarity.
A hybrid software engineering system featuring multi-model council deliberation (cloud LPUs and local models), dynamic AST code self-mutation, and isolated Docker sandboxing with zero host compromise.
Architected automated Power BI dashboards and SQL pipelines providing real-time telemetry on SLA adherence, cycle-time bottlenecks, and catalogue quality governance across high-volume ecosystems.
Designed specialized tooling for automated QA auditing, discrepancy detection, headless web scraping, and Lean process optimization that removes manual bottlenecks.
Deconstruct problems to their core physical and logical truths before assembling a scalable architecture.
If a task repeats, it should be codified, automated, and governed by deterministic metrics.
Autonomy without safety is liability. Real engineering tests in isolation before live hot deployment.
Great software feels lightweight, resilient, and effortless to whoever relies on it every day.
Rajiv Gandhi Proudyogiki Vishwavidyalaya (RGPV), Bhopal