01 · Financial intelligenceLive →
Feen.
Accountants drowned in invoices and statements. Feen extracts, structures and reconciles them automatically — shipped as web, API, and mobile.
Python · FastAPI · Next.js · Postgres
Nuva · AI Studio01 · AI systems & products
Build less.
Ship more.
An AI studio shipping agents, data platforms, and production AI products that pay back from week one. No notebooks, no demos — live systems your team actually uses.
Three partners · One studio
10+
Products shipped
4
Domains
3
Cities
Services · What we do02 · Four practices
AI agents & products
that earn their keep.
Four practices, one senior team. Pick the shape that fits the problem — every engagement ships to production, never a slide deck.
- 01
AI agents & automation
Agents that replace manual review loops — extraction, classification, triage — running unattended with guardrails and audit logs.
Claude · Python · FastAPI · Postgres
- 02
Data platforms & RAG
Knowledge bases and retrieval pipelines that turn unstructured documents into a queryable source of truth for your team.
RAG · Postgres · pgvector · Python
- 03
AI-powered web products
Full-stack Next.js apps on top of the model — dashboards, operator consoles, customer apps. Typed, localized, billable, live.
Next.js · TypeScript · Supabase · Stripe
- 04
AI advisory
Architecture reviews, build-vs-buy calls, vendor shortlists. You leave with a plan that ships, not a slide deck.
Strategy · Architecture · Review
Every engagement ships to production · No PoC cul-de-sacs
Work · Shipped03 · Selected projects
Production AI.
Real users.
A sample of the systems we've built — all live, all earning their keep. Ask about what isn't public.
02 · AI operationsLive →
Synq.
Teams lost track of what their agents were doing. Synq orchestrates multi-agent workflows with full observability — one control plane for production AI.
Python · TypeScript · Next.js
03 · Knowledge OSLive →
Pixl-KB.
Internal agents couldn't trust their sources. Pixl-KB gives them a typed knowledge graph with entity extraction — the retrieval layer behind our dashboards.
Python · RAG · pgvector
More on request · Ask about what isn't public
Approach · Thesis04 · Why Nuva
Production AI still takes too long, burns too much compute, and dies in too many proof-of-concepts.
We ship AI products that run on weekdays, handle real traffic, and close the loop to revenue. Small senior team, honest scope, boring infrastructure.
01 · Posture
Production-first
Every engagement ends in a live system — monitored, tested, documented. No PoC graveyard.
02 · Posture
Senior-only
You work directly with the people building it. No account managers, no juniors on your bill.
03 · Posture
Boring infra
Next.js, Postgres, FastAPI, Claude. Tools that will still be here in three years.
Team · Founders06 · Who builds it
Three partners.
One studio.
Small senior team. You work directly with the people writing the code, shipping the systems, and signing the contracts.
Hamza Mounir
Product & Engineering
Founding partner. Ships products, keeps the studio running, writes most of the TypeScript.
- Based in
- Brussels
- hamza@nuva.be
- View profile →
Tarek El Abdel-Laoui
Growth & Operations
Founding partner. Data & analytics background, 4+ years. Co-founder at Feen, runs sales, growth, and ops across the studio.
- Based in
- Brussels
- tarek@nuva.be
- View profile →
Sanawar Syed
AI & Systems
Founding partner. 10+ years in tech and startup engineering, multilingual, ULB alum. Builds the AI systems that go live.
- Based in
- Brussels
- sanawar@nuva.be
- View profile →
Founding partners · No juniors on your bill
Voice · Boilerplate · 30 words
“Nuva is an AI studio for companies that need outcomes, not demos. We ship agents, data platforms, and web products that close the loop between the model and the business.”
— Hamza · Founding Partner · Brussels
Commerce · Engagement05 · Three shapes
Fixed scope.
Clear outcome.
Three ways to work with us — from a one-week AI prototype to an embedded senior team. Fixed EUR pricing, every engagement shipped to production.
Prototype sprint
€8k/ 1–2 weeks
Prove one AI use case works — or doesn't — on your real data, in one week.
- ·One scoped use case
- ·Working prototype on your data
- ·Cost + latency benchmarks
- ·Clear go / no-go verdict
Product buildMost picked
€25k+/ 4–8 weeks
A shippable AI product. Fixed scope, fixed price, delivered to your production infra.
- ·Full stack: model + API + UI
- ·Fixed scope, fixed price
- ·Deployed to your infrastructure
- ·Handover docs + 30-day support
Embedded team
€18k/ mo
A senior AI team working as yours — monthly retainer, weekly shipping, ongoing outcomes.
- ·Dedicated senior engineers
- ·Weekly delivery rhythm
- ·Roadmap ownership
- ·Direct line to Hamza
Belgian VAT 21% applied where required · Brussels-based contracts
Nuva · AI Studio2026
Let's
build.
Send one line: the problem worth solving and the constraint in the way. We reply within a working day with a shape we could work in — or the name of someone better suited.
Nuva BV · Brussels