About
I'm an engineering-physics-trained systems builder. My work lives where product, operations and AI meet.
I'm an engineering physics graduate building validated CFD simulations. I work at the intersection of fluid dynamics, HPC, and GPU-accelerated computing.
Why Engineering Physics Still Matters
I spent years modeling real-world systems: dynamics, noise, uncertainty, feedback. That training shows up today in how I design AI systems—as state machines with clear boundaries, not just "prompted agents."
My engineering physics training is the foundation of everything I do in CFD. Modeling real-world systems under uncertainty, feedback loops, and nonlinear dynamics — that's exactly what computational fluid dynamics demands.
CFD & HPC Skills
Journey
From engineering physics to AI systems architecture
Foundations (Engineering Physics)
Modeled physical and control systems under uncertainty
- Designed systems with feedback loops, noise, and nonlinear dynamics
- This forced me to think in states, constraints, and tradeoffs—not just features
Q-Access: Product Without Economics
Shipped a multi-role campus access system with happy users—but no buyers
- Validated usability but never validated the problem with the budget holder
- This revealed: users ≠ buyers → now I start with the economic buyer before building
Business & AI Fundamentals
Deep study of JTBD, pricing, storytelling, and negotiation
- Built GPTs embedding business frameworks to pressure-test my thinking
- This gave me a language for why products fail beyond "bad code"
Automation Systems
Engine & Nana's Clean—real revenue workflows
- Automated intake and booking flows; Nana's Clean went from $2.2k to $5k/month
- This taught me: automation is workflow design, not tooling. AI is a component, not the system.
AI Systems in the Wild
Hemoeco & SupervisorAI—hit real production limits
- Built chatbots and voicebots; encountered infra failures and non-deterministic behavior
- This forced me to accept: control over state and cache is non-negotiable in AI systems
AI FrontDesk
Building a platform for digital employees
- Validated problem with 3 paying clients; now architecting for reliability at scale
- This is where I committed to DDD/CQRS and state machines—architecture before speed
And the journey continues...