01
Language and agentic systems
Assistants, analysts and agents that answer from your documents and act on your systems, with evaluation so every output traces to a source.
Proven on ANDRE, Hadin Co-Advisor
- RAG
- Extraction
- Agents
- MCP
- Evaluation
Ambition, translated into production
We take the idea out of the deck and the strategy out of your data, then turn both into a product that is live, measured and yours. An encoding, not a rewrite.
Language and agents
Assistants, analysts and agentic workflows grounded in your own data.
Computer vision
Detection and scoring on imagery you already collect, from phone to fixed camera.
Forecasting and decisions
Probabilities, simulations and ranked recommendations people can act on.
Product platforms
The SaaS, pipelines and cloud that turn a model into a product.
V2hhdCB3ZSBkbw== What we do
Any stage, from idea to operation
Complex platforms built from the ground up, and help at whichever phase you are in. Most clients arrive at one stage and stay through the next.
- 01
Idea
Bullet-proof it before anyone writes code.
Pressure-test the problem, the market and the data you would need. A written verdict in two weeks, including when the answer is no.
- 02
Research
Feasibility with numbers.
Data audit, model choice, a working proof on a slice of real data, and a cost to build it properly.
- 03
MVP
A real product in weeks.
The first version your users can touch, built on the architecture the full product will need.
- 04
Scale
Ready for the first big customer.
Multi-tenant, secure, compliant, observable. The work a buyer's due diligence asks about.
- 05
Operate
Keep it running and growing.
Monitoring, retraining, a team on call, or your own engineers trained to take over.
What we build
Four practices, one delivery team, each proven on a product that is live today.
02
Computer vision
Detection, scoring and capture pipelines on phone, dashcam or fixed cameras, from the capture app to the dashboard that ranks the results.
Proven on Waysights
- Detection
- Scoring
- Mobile capture
- Edge
03
Forecasting and decision models
Causal models, ensemble forecasts and Monte Carlo simulation turned into one ranked view of what to do next.
Proven on Anthos, Kumo Edge, Hadin
- Simulation
- Probabilistic forecasting
- Decision queues
- Explainability
04
Data and product platform
Multi-tenant SaaS, live data pipelines and the security work a buyer asks about: tenant isolation, audit logs, data residency.
Proven on Hadin SOC 2, Kumo live feed
- Pipelines
- SaaS
- Security
- Compliance
V29yaw== Work
MDEgLyAwNQ==01 / 05
Anthos Decision support
Make the hard decision auditable.
Causal models tied to a real decision: scenario planning across twenty countries, five hundred Monte Carlo runs on a deal structure, AI-assisted research beside expert judgment.
500simulations per decision
V2hvIHdlIGhlbHA= Who we help
Founders, teams and the people backing them
Founders with an idea
You have the domain and the problem. We bring a team that has shipped five AI products from zero, and we tell you early if the idea does not hold.
Early-stage teams
A senior build partner that works inside your repositories and your cloud, then hands over or stays on.
Investors and operators
Technical due diligence before a cheque, or a platform that needs rescuing after one.
5AI products shipped from zero
10+ yrsBuilding production systems
4Cities: Doha, Dubai, Algiers, Paris
SG93IHdlIHdvcms= How we work
Ways to work with us
Pick the one that matches where you are. Moving between them mid-engagement is normal.
Assess
A two-week read on whether the idea holds, and what it would take to build.
- Problem and market pressure-test
- Data and feasibility review
- Written verdict with a cost
Build
A dedicated team from first prototype into production, billed monthly.
- Lead engineer, ML specialist, product lead
- Fortnightly demos
- Code and models stay yours
- Handover documentation
Run
Ongoing operation of what is live, or a fractional AI team inside yours.
- Monitoring and retraining
- Agreed response times
- Quarterly roadmap review
UXVlc3Rpb25z Questions
Before you ask
We only have an idea and a deck. Is that too early?
No. That is where the assessment starts. Two weeks later you have a verdict on the idea, a view on the data you would need, and a cost to build it. Sometimes the verdict is no, and you have saved two quarters.
Who owns the models and the code?
You do, from day one. Everything is delivered into your repositories and your cloud accounts, with documentation your own engineers can pick up.
Can you work with our existing engineers?
Yes. We deliver into your repositories, demo every fortnight, and write documentation your team can pick up. Where there is no team yet, we are it until there is.
How do you handle sensitive data?
Deployment inside your environment is the default. Where models run on third-party infrastructure, we tell you exactly what leaves your network and what the retention terms are.
What does an investor get from you?
A technical due diligence read in one to two weeks: architecture, data, model claims, the team's velocity, and what it would cost to fix what we find.
Where are you based?
Doha, with teams in Dubai, Algiers and Paris. Engagements run remotely, with on-site time when it matters.
Tell us what you want to build
Thirty minutes is usually enough to tell whether there is something here. No deck, no discovery questionnaire.
- DohaQatarHeadquarters. Address to come.
- DubaiUnited Arab EmiratesOffice. Address to come.
- AlgiersAlgeriaOffice. Address to come.
- ParisFranceOffice. Address to come.