About Me
Passionate about building intelligent systems that solve real problems. With a background in engineering and AI, I focus on creating meaningful impact through technology.
Background
I educated as an Aeronautical Engineer, moved into industrial management, and found my way to building AI systems. That path taught me to think about systems—not just the technology, but how people use it, how teams adopt it, and what breaks when you scale.
Most of my work is in enterprise AI: chatbots that connect to real knowledge bases with proper permissions, agents that orchestrate tools safely, RAG pipelines that don't hallucinate because they're grounded in source documents. The kind of systems that have to be secure, auditable, and reliable—because they're running in production for teams that depend on them.
I focus on the unglamorous parts that make AI actually work: how do you chunk documents so retrieval doesn't suck? How do you handle permissions when users query across multiple systems? How do you build eval harnesses so you know if your changes make things better or worse?
I don't chase hype. I build systems that ship, get adopted, and keep working after launch.
Education
University of Brighton, UK
BEng (Hons), Aeronautical Engineering
Engineering fundamentals: systems design, optimization, failure analysis. Learned to think about how things break under real-world constraints.
University of Texas, USA
MS, Industrial Management
Bridging engineering and operations: how teams adopt technology, how to measure impact, how to build systems that scale beyond prototypes.
How I Work
Build for production from day one
Prototypes are fine for demos, but I design for what happens after the demo: permissions, error handling, audit logs, eval harnesses. The systems I build need to keep working when they scale and when things go wrong.
Measure everything that matters
If you can't measure it, you can't improve it. I build eval sets, track metrics, and run experiments before making changes. Vibes-based optimization doesn't work at scale.
Security and auditability are not optional
Enterprise AI needs guardrails: least-privilege access, permission passthrough, full audit trails. These aren't features you bolt on later—they're architectural decisions from the start.
Simplicity over cleverness
The simplest solution that works is usually the right one. Add complexity only when you have evidence it's needed—not because it might be useful someday.
Technical Stack
AI & LLM Systems
Backend & APIs
Frontend & Web
Cloud & Infrastructure
Approach
Work with Me
I'm open to full-time roles, contract projects, and advisory work. If you're building AI systems that need to work in production, let's talk.