AI agents
We design your AI agents and products, connect them to your tools and monitor how they work in production.
View this service →
Shape what comes next.
We design your AI agents and products, put them into production and operate them.
Scoping within 48h · no commitment.
Choose where to start.
Fractional CTO / CPO
Frédéric Volle leads the technology and product work on your project through a fractional or interim engagement.
View this service →Data and AI leadership
Kawtar Kostani leads your data and AI work, from team data literacy to managing projects in production.
View this service →Two engagements you can see.
Legaleo.ai
Product and technology leadership for a legal platform serving franchise networks.
Visit the website →Urbabox
Website, acquisition and management cockpit for a self-storage business in Montélimar.
Visit the website →
THE ENUMALON STANDARD
Our method.
Understand, build, operate, hand it back.
Human direction at every phase.
Before building anything, we know where AI creates value for you, and where it doesn't.
AI scoping
We map your processes and pinpoint where AI saves time or money. No tech for tech's sake.
Kheopsys · innovation offer overhaul, CEO mandate
AI-driven market study
Competition, market size, weak signals. A documented study, sources included, built in a few days.
Urbabox · self-storage market in Montélimar
Positioning & advisory
We structure the offer and product before writing a single line of code. The right problem first, the right solution next.
Cartalys · product structuring, CPO role
Data governance
Mapping, quality, lineage, GDPR and AI Act compliance. Your data under control, not in the wild, with who-can-do-what written down.
Yavantha · Head of Data & AI, 3 CDO mandates
You leave with
A costed scoping note. Audit from EUR 2,000, credited against the project if you continue. Go or no-go, no commitment.
We don't deliver a slide or a mockup. We deliver a product that runs, online, used by your clients.
CTO / CPO as a service
A technical AND product lead, rented for the time you need. One head for architecture and product: zero friction between them, no six-figure hire.
CPTO Legaleo.ai, incubated at the Paris Bar · CPO Cartalys
AI product in production
From the agent workshop to the live product. We build for real, with tests and monitoring, not a POC asleep in a corner.
Our workshop · 42 agents → Legaleo, OverTh1nk, d3rf.com
Site & integrations
A site readable by humans AND by AI (llms.txt, GEO, MCP), connected to your existing tools. The new storefront is also the AI's answer.
enumalon.com · this site applies what it sells
Anomaly detection
Engines that spot fraud, errors or abnormal behaviour in a real data flow. Applied machine learning, not a lab demo.
Swapli · fintech fraud detection · swapli.ma
Turnkey SME offer
For an SME: a simple scope, a clear price, a product that runs with no tech team on your side. The same foundation moves from one sector to another.
Urbabox · self-storage, transposable model
You leave with
A live product, tested and monitored. First visible result within 7 days.
A delivered product is not a living product. We pilot it, we measure it, and it improves on its own, loop after loop.
4 820Visitors · 30d
37Bookings
21 €Cost / booking
2,8xAd return
780 €Ad budget
0,8 %Booking rate
Best day
Bookings · 30d
In plain words
For EUR 780 of advertising this month, you got 37 bookings, at EUR 21 each. Friday remains your best day.
Operations cockpit
A dashboard showing the product's real state: traffic, usage, business metrics. A wrong number is worse than no number, so we show the truth.
Urbabox · real-time cockpit, GA4 + Plausible, in production
Live advisory assistant
An AI that answers, guides and explains continuously on your own data. Not a scripted chatbot: an assistant that knows your context.
Our agent · live on this page
Presence inside the AIs
We measure whether ChatGPT, Claude or Gemini cite your brand when asked. Tomorrow, being found means being cited by AI, not only by Google.
GEO · citation tracking measured, not promised
Self-improvement loops
After agents come loops. An agent executes; a loop learns from what it just did, then starts again better. The system re-reads, corrects and remembers itself.
Our workshop · ~146,000-entry memory + living agent map
You leave with
A real-time dashboard. You see the system's state, not a monthly report.
The goal is not to make you dependent. It's to hand it back: your teams know how, and everything is yours.
Upskilling & training
We train your teams to pilot AI themselves, from the boardroom to the floor. Twenty years spent translating tech for non-technical people.
Enumalon Academy · Executive Pack, 10 sessions, from '70 years of AI in 2h' to 'brief an agent like a junior'
Documentation & ownership
Code, accounts, access, infrastructure: all documented, all yours. Open-source stack first, zero vendor lock-in imposed.
Open stack · no vendor lock-in
Talks & culture
Talks to bring leaders and teams on board, and make AI understandable rather than magical. We speak complexity without hiding it.
TEDxENSAMrabat 2026 · 'Embrace Complexity' · 20 years Apple → Ledger
You leave with
Full ownership. No dependency on Enumalon imposed by default.
Two careers before the agents.
Twenty years of product and technology on one side, the data of major companies on the other. The agents learned from us.
Houses that know us.
Demonstration · example data
THE COCKPIT // LIVEMeasured, explained, decided. Live.
Your output measured, explained by AI, and the actions that follow. Live, in the language of each role.
FOUR ROLES · FOUR SETS OF METRICS
Incoming requests · 30 days
Applications · 30 days
Invoices received · 30 days
Work orders · 30 days
2,4804401,2301,940this month
target 90 / daytarget 12 / daythreshold 70 / daytarget rate 60 / day
D-30the 12ththe 12ththe 26ththe 14thtoday
AI reading
108 requests a day since the 12th, against 38 before. They all come through the training page.
Action: double the budget of the ad feeding it.
20 applications a day since the 12th, against 6 before. The ad posted on the 11th does all the work.
Action: cut the other two channels and keep this one.
108 invoices on the 29th, against 31 on ordinary days. Month end carries a third of the volume.
Action: put two people on the last five days.
Three days below target from the 14th to the 16th, line 2 down. Caught up in four days.
Action: schedule the next maintenance in the week of the 20th.
Quotes by weekday
Applications by weekday
Documents received by weekday
Checks by weekday
MON
TUE
WED
THU
FRI
SAT
SUN
AI reading
Thursday carries twice as many quotes as Monday.
Action: block Thursday morning for drafting.
Sunday brings in as many applications as Wednesday.
Action: review the weekend batch on Monday morning.
Friday carries a quarter of the documents, the weekend almost none.
Action: drop weekend monitoring.
The load rises 60% from Tuesday to Thursday, then falls back.
Action: move part of the checks to Tuesday.
Quotes followed up on time
Complete applications
Invoices matched automatically
Parts within spec
94%88%97%91%
6% to redo12% to complete3% to review9% to redo
AI reading
94% of quotes are followed up on time. The 6% missed have been sitting for over eight days.
Action: follow them up today, oldest first.
88% of applications arrive complete. The incomplete 12% come from the mobile form.
Action: fix the field that drops on phones.
97% of invoices match automatically. The remaining 3% come from the same supplier.
Action: fix that supplier record once and for all.
91% of parts are within spec. Every reject comes from the same station.
Action: set up machine 3 before the next run.
Cockpit demo · illustrative data (client confidentiality)
AN ENUMALON AGENT // LIVE
Ask your question.
An agent trained on our methods, our products and our values. We can build yours from your documentation, your experience, your clients.
TRAINING // HANDING OVER
We help you take back the controls.
A four-chapter path, from leadership to the field.
Our products.
We built them. We are opening them one by one.
sellom.io
SOONYour customers no longer search on Google: they ask an AI. A product, a solution, a provider? Yours is the one it should suggest. sellom.io measures whether ChatGPT, Claude and Gemini cite you, and how you rank in your field, against your competitors.
Opens September 10, 2026
--days
--hours
--min
--sec
-
Skopalon
SOONThe cockpit for your market, your customers and your campaigns. AI analysts read the numbers and recommend your next moves.
OPENING // SOON -
OverTh1nk
SOONYour ideas are the seed, OverTh1nk makes them grow. Expanded from every relevant angle, checked, kept, enriched, made actionable for you.
OPENING // SOON -
AL
LIVE INSTANCEThis site's assistant. It answers on our methods, live. Adaptable to your market, your clients, your documentation, your habits, your KPIs, your goals.
PRODUCT OFFERING COMING SOONLive instance on this website. Product offering coming soon.
Try AL -
Prodium
SOONThe agent workshop that builds our products and learns from every project.
OPENING // SOON
Ergalon
IN DEVELOPMENTIt measures your company's real AI usage: costs, carbon, attestations.
WORKSHOP // IN PROGRESSDemos and prototypes
How we build our tools.
-
ToutVaMal
demoA satirical outlet written and illustrated on its own, three times a day.
What it taught ussee the demoA chain that publishes three times a day with nobody at the controls, cost tracked per article.
-
GPSwitch
liveA GPS file converter, in production, used every day.
What it taught ussee the demoA public converter kept in production: the code follows real files.
-
SpotHuman
RDA public test: can you tell a human text from one written by an AI?
What it taught ussee the demoA public test that measures where readers think they spot an AI, and where they get it wrong.
-
DataSpy
demoA public scan showing what a website learns about you in thirty seconds.
What it taught ussee the demoShowing someone what their browser reveals convinces faster than ten pages of explanation.
Let’s discuss your project.
An exploratory call to understand your project. The scope, deliverables and proposal are agreed together.
Explore our workshop and services
What your agents will do.
These examples combine. They exist to pinpoint your needs, not to tick a list.
The partner agent
- Knows your role and your company.
- Sees only what you are allowed to see.
- Backs your decisions and flags what is urgent.
- Writes your emails and reports in your voice.
- Briefs a colleague while you are away.
The brand agent
- Speaks for your brand, in your tone.
- Answers your clients at any hour.
- Qualifies requests before troubling a human.
- Hands over when a question calls for it.
The intake agent
- Greets every incoming request.
- Grasps the real need behind the ask.
- Routes to the right person, the right team.
- No one waits anymore.
The watch agent
- Reads the market, rivals and regulation, non-stop.
- Reports only what changes your decisions.
- Alerts when a weak signal turns serious.
The critic agent
- Reviews the other agents' work.
- Checks facts and figures before release.
- He has the right to say no.
- He is the one who holds quality.
The follow-up agent
- Tracks every quote and proposal you send.
- Follows up at the right time, in the right tone, never forgets.
- Prepares your meetings: history, context, open points.
- Flags what deserves your call rather than an email.
- No quote sleeps.
THE SYNTHESIS // THE ORGANISATION
An organisation of agents.
Narrow specialists, one orchestrator, control loops.
That is the Enumalon product: we design the organisation that looks like you.
What the agents taught us themselves.
'A strong agent is a narrow agent.'
A precise role, a refusal criterion, a right of veto. The sharper the scope, the safer the work.
'Skills are earned in a loop.'
Every corrected mistake becomes a rule. When agents review one another, skills no one programmed appear.
Pressure breaks agents.
A system driven by urgency takes shortcuts and ships the false, like a team marched too hard. Quality comes from calm organisations, with controls.
THE PROOF
Products in production.
THE LOOP // NOT THE PROJECT
Not a frozen deliverable. A loop that learns.
An agent delivers once · a loop measures itself, checks itself and sharpens with every iteration.
The workshop keeps the loop turning
- Deliver A product that runs, online.
- Measure Real numbers, not a hunch.
- Verify The system checks its own work.
- Refine Each iteration gains precision.
The principle is simple: the system checks its own work before handing it over, corrects itself, then starts again better.
In its internal research evaluation, Anthropic reports a 90.2% performance improvement for a multi-agent system compared with Claude Opus 4 alone.
Well orchestrated, several viewpoints beat a single one, and our workshop keeps this loop turning project after project.
01 // CTO · CPO AS A SERVICE
CTO / CPO as a service
by Frédéric Volle
Co-founder · Chief Product & Technology Officer
One head for tech, product and adoption. Zero friction.
see the detail
collapse
CPTO: technical and product leadership in one person.
A shared part-time tech and product executive, no hiring.
01 Technical Due Diligence Assess technology before it becomes a negotiation sticking point.
Thorough review of the tech stack, architecture, and technical debt ahead of fundraising or M&A. Assessment of code, security, scalability, and team structure, delivered in actionable terms for investors or buyers. The goal: safeguard valuation, not just tick boxes.
Ledger · technical input on merger and acquisition files and enterprise blockchain integration.
02 Turnaround Leadership Regain control when delivery fails to meet commitments.
Rapid diagnosis of bottlenecks, takeover of product and technical leadership, and restoration of predictable delivery. Alignment of product, engineering, and sales teams to re-establish a reliable trajectory.
Proximus · an unstable release flow turned into a predictable cadence on a B2B SaaS platform.
03 Embedded AI Strategy Drive AI strategy within your organization, not alongside it.
Define and execute AI strategy with existing teams, without replacement. Prioritize use cases, select architecture, arbitrate build vs buy, and implement through to production.
Kheopsys · direct advisory to the CEO, three projects delivered, company repositioned as an AI partner.
04 Technical Debt Audit Determine what to fix, discard, or acquire before investing.
Review of architecture, code, and infrastructure choices. Risk mapping, build vs buy recommendations, and remediation roadmap prioritized by business impact.
Scality · Publicis · petabyte-scale data architectures · data-centre consolidation under pressure.
05 AI System Compliance Document and secure AI systems ahead of full AI Act enforcement.
Technical compliance layer: risk classification, documentation, human oversight, and traceability of AI systems (models, agents, pipelines). Complements data governance without overlap.
Ledger Vault · institutional KYC/KYB onboarding on one of the most regulated custody platforms on the market.
06 Interim CTO/CPO Hold the role, structure the function, and prepare the handover.
Immediate assumption of the role while recruiting a permanent hire. Structuring of technical and product functions, establishment of rituals, visible early results, and handover playbook.
Legaleo · Swapli · two fractional mandates held in parallel today.
07 0-to-1 Product Leadership Build a platform or marketplace from the first line of code.
CPTO role for a non-existent product: architecture, team building, early user acquisition, and adoption metrics. Technical and product bootstrapping through first releases.
Swapli · the first ticket resale marketplace in Morocco, built from zero.
02 // CDO · CAIO AS A SERVICE
CDO / CAIO as a service
by Kawtar Kostani
Co-founder · Chief Data & AI Officer
Seven fields, one data leadership. From culture to measurable ROI.
see the detail
collapse
CDO and CAIO: data strategy and AI strategy, together.
A shared, part-time data leadership, from literacy to production.
01 Data & AI literacy Translate technical stakes into business language and align the executive committee.
A data and AI literacy programme for the executive committee and management: translating technical stakes into business language, strategic prioritisation of initiatives and alignment of decision-makers around a shared vision. Workshops, awareness sessions and change support to anchor a genuine data culture at the highest level.
02 Data diagnostic and roadmap Audit maturity, target quick wins, deliver an actionable roadmap.
Data maturity audit: culture, tools, uses, organisation. Identification of priority use cases and quick wins. A costed, sequenced and actionable roadmap, then delivery oversight through to the first measurable results.
03 Data teams as agile squads Design the operating model and launch cross-functional squads working agile.
Design of the data operating model: roles, RACI, delivery. Cross-functional squads (data engineering, data science, business) working in agile mode. Governance frameworks, rituals and skills growth across the teams.
04 Governance and GDPR compliance Frame roles, quality and data assets, and secure GDPR compliance.
Governance framework: roles, data owners, committees. Management and valuation of the data estate: catalogue, dictionary, lineage. End-to-end quality: rules, controls, indicators and remediation. GDPR compliance: records of processing, consent management, retention policy, security and respect for individual rights. Governance in the service of trust and risk control.
05 Data science and data marketing Segment, score appetite and churn, and turn models into ROI.
Actionable customer segmentation, appetite scoring to target campaigns, churn scoring to anticipate attrition. Models translated into concrete marketing levers and measurable ROI.
06 AI-augmented BI and reporting Next-generation dashboards and a chatbot that queries data in plain language.
Next-generation dashboards and reporting, enriched by AI. A data-analyst chatbot to query data in natural language and democratise access to insight across the organisation.
07 Data project leadership Frame, prioritise and steer the data programme through to production.
End-to-end data project and programme leadership: framing, backlog structuring, prioritisation and delivery. Budget, planning, risk and stakeholder management. Coordination of technical and business teams, running the governance bodies (steering and technical committees) and progress reporting through to production.
Data governance
The groundwork that makes your data usable. We lay it with Yavantha, our governance partner.
Data mapping
Compliance & GDPR
Quality & traceability
Access & security
Lifecycle
03 // PARTENAIRES
They build with us.
Predictive threat intelligence: mapping adversary infrastructure to anticipate attacks before they launch.
cartalys.com
The legal platform for franchise networks. Product and technology leadership by Enumalon.
legaleo.ai
Ticket resale marketplace in Morocco. Startup support and CPTO leadership by Enumalon.
swapli.ma
City-centre self-storage in Montélimar. Website, acquisition and steering cockpit by Enumalon.
urbabox.fr
Data consulting and support. Innovation offer redesign with Enumalon, an executive mandate.
kheopsys-technology.comOne name, two origins.
Enūma
𒂊𒉡𒈠
Akkadian: “When...”
The beginning
The first word of the Enūma Eliš, the Babylonian creation epic.
Mellon
μέλλον
Ancient Greek: “The future”
What is to come
From the verb μέλλω (mellō): to be about to. What announces itself, what is to come.
When the future begins. Every client project starts from this question: what is possible now, and what will be possible in 6 months?