Hire · freelance & consulting
Hire an AI developer
who actually ships.
LLM agents, RAG pipelines, and the full-stack product around them — built end-to-end by one engineer with real systems in production, not a slide deck. Remote from Agadir, Morocco.
Six worlds shipped — hover to enter one
16.
products live in production — four of them my own
58.4K.
search impressions in 28 days — FASL, zero ad spend
818.
clients registered in Belmo's first two months
6+.
years shipping software with real users and real money
The work
What you can hire me for.
- 01
AI agents & multi-agent pipelines
FASL · RESO Khdma
Tool-using agents and staged pipelines with structured-output contracts, grounding, and human-in-the-loop review — the architecture running FASL in production.
- 02
RAG over your own data
FASL — grounded in Moroccan law
Retrieval pipelines over documents, legal codes, catalogs, or knowledge bases — so the model cites your facts, not its imagination.
- 03
Full-stack AI products, zero to launch
Belmo · Magical Hekaya
Next.js / TypeScript / PostgreSQL from empty repo to paying customers: payments, auth, admin, analytics, deployment on Vercel or your own server.
- 04
Arabic-first & multilingual builds
4 languages · RTL-first
RTL-native interfaces and models that actually understand Arabic, Darija, and French — for the markets most agencies get wrong.
The receipts
The proof is running in production.
The method
How an engagement works.
01
Scope
A short call or email thread: your problem, your data, what “working” means. You get a concrete plan and quote.
02
Build
Short iterations with something clickable early. You see progress weekly, not at the end.
03
Ship
Deployed on Vercel or your own infrastructure, with payments, analytics, and admin where needed.
04
Iterate
Documentation, credentials, and a codebase your team can own. No lock-in, no mystery boxes.
The questions
Asked by every client.
If yours isn't here, the letter at the bottom of the homepage takes about forty seconds.
What kind of AI developer are you?
A full-stack engineer specialized in LLM-powered products: AI agents with tool use, multi-agent pipelines, RAG over domain documents, and the product around them — Next.js, TypeScript, PostgreSQL. I build the whole thing, not just the model calls.
Can I hire you to build an AI agent?
Yes — that's the core of what I do. I've built production agent systems: a multi-agent legal drafting pipeline (FASL), a WhatsApp job-matching agent that interviews workers in Darija (RESO Khdma), and an automated video-ad generation pipeline (Lakta).
Do you work with small businesses or only startups?
Both. Small businesses usually need one well-scoped AI capability wired into how they already work — a support agent, document automation, search over their own data. I scope small, ship fast, and you own everything.
How do you charge — freelance projects or ongoing consulting?
Both models work: fixed-scope projects (an agent, a pipeline, an MVP) or an ongoing consulting/engineering arrangement. Every engagement starts with a short scoping conversation so the quote reflects the actual problem.
Do you build Arabic-first or multilingual AI products?
Natively. I'm an Arabic and Darija native speaker; FASL runs Arabic-first with full RTL, and RESO Khdma understands Arabic, Darija, and French. If your market speaks Arabic or French, it's built for it — not translated as an afterthought.
Where are you based, and do you work remotely?
Agadir, Morocco (GMT+1 — European hours overlap). I work remotely with clients anywhere, and everything ships with documentation, deployment, and a clean handover.
One engineer · the whole lifecycle
Salam — tell me what
you're building.
One honest paragraph about your problem is enough — I reply to everything myself, within 24 hours.
Start the conversation →