About
I build things that have to be correct, not just impressive
A software engineer specializing in backend development, and increasingly, applied AI and machine learning. I like the parts of engineering that are unglamorous but consequential: the data model, the auth boundary, the API contract.
Engineering philosophy
I think of software as a set of promises: an API promises a certain response shape, a database schema promises certain constraints hold, an auth layer promises certain data stays scoped. Most serious bugs I’ve hit a student’s progress data leaking across accounts, a silent audio-upload failure came from a promise that was assumed rather than enforced.
So I build defensively at the boundaries: validate inputs, validate model outputs, scope queries explicitly, and test the cases that only show up under realistic, concurrent use rather than a single happy-path click-through.
Approach to scale and AI
Scalable software, to me, starts with a data model that won’t need to be re-architected the moment a second user shows up which is a lower bar than it sounds, and one that’s easy to miss under deadline pressure.
AI and machine learning interest me specifically as a layer on top of that foundation not a replacement for good backend design, but a way to extract structure and insight from messy, real-world input once the underlying system can be trusted to handle it correctly.
My journey
Foundations
From operations to systems thinking
A background in sales and operations gave me a close-up view of how disconnected data and manual processes slow real organizations down the same pattern that later became the starting point for Dataloom.
Training
Software engineering & QA
Structured, project-based training covering full-stack development, API design, SQL, and QA automation working through problem sets and building real applications rather than isolated exercises.
Building
Shipping full-stack and AI-integrated projects
Built and debugged production-shaped systems: a FastAPI + Next.js progress tracker, an AI-assisted health records platform, and automated test suites learning most from the bugs that only show up under real, concurrent use.
Founding
Dataloom and the data-silos problem
Researching and building Dataloom, a DaaS platform addressing data fragmentation among East African SMEs currently in active development.
What I bring
Strengths I lead with
- A habit of tracing bugs to their actual root cause a missing filter, a wrong response shape instead of patching symptoms
- Comfort moving between backend, frontend, and AI integration on the same project without losing coherence
- A research-first instinct: before building, I want evidence the problem is real and specific
- Direct, honest communication about project stage I won't tell you something is production-ready when it's a prototype
Where I'm headed
Career goals
I'm working toward roles and projects where backend engineering and applied AI meet directly building systems that use AI responsibly, are tested and secured properly, and solve problems I can clearly explain the value of. Dataloom is my attempt to build that at company scale; I bring the same standard to every project, prototype or production.