Behind the systems, one point of execution.
As an AI researcher and systems architect, I build software that bridges neural models with physical reality. Direct technical execution, zero boilerplate, and uncompromising precision.
From curiosity to autonomous systems.
I am Muhammed Arshad, an independent AI researcher and builder based in Kerala, India. My work focuses on constructing autonomous multi-agent swarms, real-time spatial computer vision pipelines, and proactive cybersecurity triaging engines.
Rather than relying on fragile abstractions or superficial wrapper scripts, I investigate systems down to their memory layouts, raw serial bus packets, and network protocol definitions.
Statement
True engineering is not about following noise. When you inspect systems at the byte level, you discover how intelligence truly scales. noise. intelligence scales.
Hands on the metal, abstracting nothing.
I operate with a systems-first mindset. When implementing agentic AI or computer vision, understanding the hardware limitations, serial latency, and cache hierarchy is what differentiates a fragile demo from a mission-critical tool.
Whether it's writing ESP32 firmware for ultrasonic transducers, deploying vLLM inference microservices on Linux GPU clusters, or profiling WebGL fragment shaders, I prefer direct ownership of the full technical stack.
How I engineer.
Every engagement follows a rigorous cycle: reverse-engineering existing barriers, isolating the mathematical or structural bottlenecks, prototyping minimal deterministic proof-of-concepts, and hardening for production deployment.
No unnecessary dependencies. No bloated libraries. Strictly performant, maintainable, and verified code.
My trajectory, in key milestones.
- Autonomous Agent & Systems Architect
- Spatial Computer Vision & Hardware Developer
- Vulnerability Intelligence & Triage Research
- Knowledge Graph Synthesis & RAG Architecture
“I don’t chase the future. I reverse-engineer it.” reverse-engineer