I'm an Applied AI Engineer, though I started out doing full-stack work, so I've shipped enough software to know the difference between a demo and something people actually rely on.
One half of my job is helping businesses automate the parts of their operation that eat time: a WhatsApp receptionist that handles hotel bookings, voice agents that chase leads, and the CRM workflow I'm currently leading development on at Globastaff that moves candidates through an entire immigration pipeline without anyone touching a spreadsheet.
The other half is helping products add AI where it's missing: RAG search, a scoring engine inside a resume analyzer, the kind of feature a client's app didn't have until I built it. Same underlying problem either way, just get the model to do something useful once real data and real users are involved.
I still write plenty of code by hand, which is exactly why I trust coding agents like Cursor, Claude Code, and Codex for the repetitive parts, plus Antigravity when I'm deep in UI work. Good tools just mean I ship more in a day.
A fast, mobile-first lead site that ranks cleanly, wires every inquiry into your CRM, and is structured so people and AI tools can find what you offer.
5 to 8 page lead-focused site on Next.js
Mobile-first layout with clear primary CTAs
Forms, click-to-call, and maps wired to CRM
Technical SEO: titles, meta, sitemap, robots, Core Web Vitals
Structured data (JSON-LD) on the pages that need it
Answer-ready page structure so AI search can cite you cleanly
llms.txt-style discovery so AI tools find your key pages
Free discovery, then a written scope and fixed quote. You own the repos and accounts. I turn down hourly staff-aug, free pilots, and work that needs a model to be certain.
Voice agent that books, reschedules, and cancels appointments around the clock.
Vapin8nGoogle Calendar+1View
Hackathons
I like building things
I actively participate in online hackathons and competitions on platforms like Devpost and Kaggle. It is incredible to collaborate with developers from around the world and build impactful AI solutions in just a few days.
Google DeepMind - Vibe Code with Gemini 3 Pro
Active/Remote
Built LabSnap AI, a smart assistant for science labs. It uses Google's Gemini Vision AI to read chemical labels and keep track of inventory automatically. I added voice controls so scientists can use it hands-free, and an AI safety system that warns them if they accidentally put dangerous chemicals together (like storing fire hazards next to oxidizers).
Contact
Available for select projects
Let's build something together
Start with how I work: scope, process, and what I don't take on. If it fits, book the call.