đź§  About

Senior Software Engineer, Roboticist, AI Practitioner, Relentless Optimizer.

I’m a senior engineer at Microsoft and a Master’s student in Artificial Intelligence. Professionally, I’ve led front-end and developer tooling efforts for large-scale platforms at Microsoft, Amazon, and early-stage startups. My focus has often been on making complex systems usable—whether that’s AI infrastructure, internal dev tools, or customer-facing platforms.

Outside of work, I’m building up deep hands-on experience in AI through my Master’s program and side projects—especially around reinforcement learning, multimodal models, and MCP-based developer workflows. My recent focus is on robotics, AI system integration, and the messy intersection of hardware and intelligence.

This site is where I document the systems I build: robots, training pipelines, control infrastructure, and the occasional teardown or build log. It’s technical, iterative, sometimes chaotic—and always hands-on.


đź”§ Current Projects

  • 🦾 Modular Cobot Platform
    A 3D-printed, quasi-direct-drive collaborative robot arm using TMC5160 SPI drivers, home-built cycloidal gears, and custom PCBs.
    Goal: Payload-class modular arms and mobile bases with swappable connectors, smart drivers, and homegrown control logic.

  • ⚙️ MLOps & Training Infrastructure
    Local and cloud job orchestration using W&B agents, AWS spot instances, Terraform IaC, Docker image optimization, and artifact management.
    Goal: Reproducible, efficient RL and multimodal training across desktop and cloud GPUs.

  • đź§© MCP Server Workflows & Toolchains
    Custom-built Model Context Protocol (MCP) server in TypeScript for orchestrating tools with GitHub Copilot.
    Goal: Chain structured tool responses, extend dev workflows, and experiment with intelligent agent interfaces.


đź§© Interests

  • Reinforcement Learning – A2C, PPO, PER, curriculum learning
  • MCP Servers & Toolchain Orchestration – Structured agent workflows with LLMs and GitHub Copilot
  • LLM Fine-Tuning & Multimodal AI – Extending open models for real-world applications
  • DIY Robotics – Cycloidal gears, custom arms, modular mobility platforms
  • Embedded Electronics – Stepper motor drivers, SPI/I2C control, sensor interfacing
  • 3D Printing – Rapid prototyping, actuator design, printing practical everyday objects
  • Web Dev Tooling – TypeScript, pnpm, Rush, webpack, and the dark art of taming dependency graphs and lockfiles
  • MLOps – Infrastructure as code, launch agents, reproducibility at scale
  • Parenting, Education & Suburban Farming – Teaching autonomy, feedback loops, and iteration through hands-on living

📚 Writings

What you’ll find here:

  • Build logs and hardware schematics
  • Model training experiments (and mistakes)
  • Infrastructure notes and orchestrator configs
  • Emerging thoughts on education, tools, and engineering culture

This site is a living project. If something’s half-baked, that’s probably intentional.