I build the half ofsoftware you don’t see.

I work on the parts of a product that nobody photographs for the landing page — the APIs, queues, schemas, and cloud infrastructure that quietly carry the load when traffic actually shows up. Four years across fintech, AI, and enterprise SaaS taught me to think in latency, throughput, and the long tail of edge cases that turn a working demo into a working product.

My instinct is to start every system with a hard question. Will this hold up under real load? Can this schema evolve without downtime? What happens when the third party we depend on rate-limits us at 3am? I design backends as if I’ll be the one paged about them — which I usually am — and that bias shapes every decision from data modeling to deploy pipelines.

Lately the work I find most interesting sits at the seam between traditional backend engineering and applied AI: wiring language models into production systems where reliability still matters more than novelty. The hard part isn’t whether the model can do the task — it’s whether the system around it fails gracefully when the model can’t.

That instinct is also why I ship my own products. Under WebCloudNest — the product studio I founded — four Android apps and games are in closed testing on Google Play right now: an offline Qur’an reader, a cinematic 3D chess game, an app locker, and an educational runner for kids. Nothing sharpens engineering judgment like being your own client.