For years, I built things for other people’s companies. Dozens of clients, different industries, constant context switching. I was good at it, I enjoyed it. But at some point I noticed a pattern: I’d build something, hand it over, and never see what happened next. Did it help the business grow? Did users love it? Did it even survive past launch?
I realized I wanted to build something I could see evolve. Not just deliver and move on, but stay, iterate, and watch the product grow over months and years. That’s what brought me to BlackPixel AI.
Before joining, I used AI tools daily — Copilot, ChatGPT, various APIs. I thought I understood how AI products work. I was wrong. Using AI is easy, but building a product around AI is a completely different challenge. The model gives you 85% accuracy; the product needs to feel like 99%. That gap between what the AI actually does and what the user expects is where most of the engineering happens.
What I’m focused on at BlackPixel: closing the gap between «the AI generated something» and «the print shop ships a finished album the customer is happy with.» That gap is full of edge cases — bad source photos, mismatched bleeds, color profiles that look fine on screen and broken in print. Most of our engineering is the unglamorous work of making AI feel deterministic in a domain where it isn’t.
I oversee engineering and technical architecture for an AI-powered photo book creation platform serving print companies worldwide. The work spans the full stack:
Before BlackPixel, 6 years across SaaS, e-commerce and media industries as a full-stack developer — building scalable backend systems, API integrations, and responsive web applications for businesses ranging from small startups to mid-size companies.
Especially the unglamorous parts — making AI deterministic in domains where it isn’t, building tools that print shops actually use day-to-day, dealing with edge cases at scale. Drop a note or request a demo.
Request a demo