Enterprise Software · AI Engineering · Cloud Platforms
We design, build and operate software, AI and cloud platforms for Government departments, State agencies and Irish companies, and we connect them to the systems you already rely on.
Nine nodes in a three by three grid, joined by one line that traces the FromNine mark. The line starts at the business, passes through data and models, software, cloud and day-to-day operations, and ends at a working system in production.
01 Size and principles
We concentrate on AI, cloud and current engineering practice, with the delivery capacity of a large firm behind it. In practice, we can put every discipline a programme needs on one team, and the people who design your system are still there when it goes live and afterwards.
2,400+people who design, build and operate systems for the public service and for private companies.
600+government projects for public bodies in EU Member States, completed or still in progress.
400+projects for private companies
Our work for the public serviceFits the systems you run
We build on the platforms, data and processes you have today, and only replace the parts that genuinely need it.
Easy to maintain for years
Each release is tested and documented before handover, so your own staff can keep changing it long after the project ends.
A person decides where it counts
Steps with real consequences wait for a person to approve them, and every automated action is logged for later review.
plus public bodies in many EU Member States
02 Services
AI only pays off when everything around it works: software people actually use, data you can trust, integrations that don’t break and the hosting to keep it all running. We deliver those pieces together, so they fit, and a single team is responsible for the whole.
Generative AI that copes with real users: based on your own data, tested before it goes live and monitored once it does.
Agents that stay inside the permissions you set and check with a person before any step that matters.
Well-governed data that your decisions and your AI can depend on.
SaaS products, portals and internal platforms that stay simple to adapt.
A cloud platform where shipping a release is ordinary and every cost is in plain view.
Systems that pass data to each other reliably, even when one of them goes down.
Clear technical advice before you commit: the choices, what each one costs you, and a roadmap you can stand over.
03 Solutions
Begin with the problem, not the model. Every example below is laid out the same way: what slows your staff down today, what we would put in place, the gain you can measure, and the points where a person makes the call.
04 Public service
600+projects for government and public bodies in the EU, finished or under way.
Government departments, State agencies and local authorities need technology that is accessible, secure and open to scrutiny, and that lasts longer than a single tender competition. Public-sector work has taught us to deliver within procurement rules, bilingual and multilingual requirements and strict data protection.
AI gets the same discipline: decisions written down, a person in control, and design that reflects the EU AI Act and Ireland’s Regulation of Artificial Intelligence Act 2026 from day one.
A dot map of Europe with the 27 EU Member States highlighted and the rest of Europe in a lighter shade. The Antwerp and Ghent offices are marked.
05 Enterprise platforms
Your customer, finance, supply chain and operational data already sits in SAP, Salesforce or Odoo. We set up, extend and connect these platforms, and link AI to them with the correct access rights, so automation happens inside the processes your staff use every day.
Summit Partner
Our work
Where AI helps
Help with case handling, with drafts that staff check before anything is sent
Gold Partner
Our work
Where AI helps
Incoming documents fed into SAP processes, with exceptions sent to staff
Gold Partner
Our work
Where AI helps
Back-office tasks automated on your Odoo data
SAP, Salesforce and Odoo are registered trademarks held by their respective owners.
06 Engineering detail
The model is just one part. The integrations, access rights, testing, approval steps and operations around it decide if it holds up at nine on a Monday, with real users on it. That is where AI projects either succeed or fade out without anyone noticing.
AI is worth only as much as the systems it can read and update. So we design the integrations first: field mapping, retries, de-duplication and errors your staff can see.
It only returns what the person asking is entitled to see. Access is enforced in the retrieval layer itself and never left to instructions in a prompt.
Answers and actions are checked against agreed criteria before release and after it. Any action with real consequences waits until a person approves it.
Infrastructure as code, separate environments and monitoring of quality, response time and cost, which lets your team operate, audit and improve the system once it is live.
Six layers are stacked in the diagram. At the top are agents and process flows, with an approval point for a person before any action with consequences. Below come models and testing, then data and search, then integration, then the core business systems that hold the records: CRM, ERP, case handling and document libraries. A cloud platform with monitoring supports them all from below.
Access and identity cut through every layer vertically, so roles, identity and the audit trail are checked at each step. A request comes in at the top and travels down through all six layers to the business systems.
Four numbered notes: integration before prompts (integration and core business systems); search that honours access rights (data and search, access and identity); testing and sign-off points (models and agents); run, not just deployed (cloud platform and monitoring).
07 How we work
Four stages. Each one ends with something you can check and keep. You see where things stand at every point, and what we build belongs to you.
We look at the business problem, who is involved and which systems are already there, then agree with you what to tackle first.
ResultA clear problem statement
We pick an architecture, agree how we will judge success and plan a first release that is small enough to go live.
ResultArchitecture and measures of success
We work in short, tested steps that your staff can see, try out and question.
ResultA tested build
We hand the system over or run it ourselves, watch it in production and keep improving it with you.
ResultHandover and a plan for operations
What we learn in production shapes the next round of discovery.
08 Trust, security and sovereignty
Responsible AI is engineering work, not a policy on a shelf. Four layers of control surround each system we build, with your data and your decisions in the middle.
Data hosted in the EU by design. We settle the hosting locations with you before building starts, including sovereign cloud options where your requirements need them.
Our information security management system is certified to ISO/IEC 27001. Every project includes least-privilege access, threat modelling and secure engineering from the design stage on.
Audit logs, decision records, documentation and risk classification, designed around the EU AI Act, the Irish Act of 2026 and GDPR, so you can explain any decision and who made it.
Approval steps, escalation routes and the power to override, wherever an automated decision really affects people.
Your data, your decisions
Credentials
The systems we build are designed for GDPR, the AI Act and NIS2, including Ireland’s National Cyber Security Bill as it takes shape.
Security and data handling09 In Ireland
Irish buyers ask three things early: who will do the work, what the system will cost to run, and how a supplier from outside Ireland stays close. Fair questions. We do not have an office in Ireland. Our teams work from Antwerp, Ghent and Amsterdam, one hour ahead of Irish time, and the people who design your system are the people who deliver it.
Much of the Irish economy is made up of multinationals with Irish operations, where decisions are often shared with a group function abroad, and indigenous Irish companies growing from home. Both need choices written down so that everyone involved can follow them. Three changes shape the agenda right now: the AI Act already has an Irish Act and an AI Office, NIS2 is still making its way through the National Cyber Security Bill, and Revenue’s VAT modernisation plans mandatory B2B e-invoicing for large corporates from 1 November 2028.
Public work taught us to build for accountability: 600+ government projects for public bodies in EU Member States, completed and under way. In Ireland that means the principles of Connecting Government 2030 (digital by default, cloud-first, secure by design), sign-in with MyGovID, accessible services under S.I. No. 358 of 2020, Irish-language duties under the Official Languages Act, and EN 16931 e-invoices over Peppol for contracting authorities under S.I. No. 258 of 2019. Opportunities appear on eTenders, and we answer each requirement of the request for tender.
A county council wants residents to apply for services online in Irish or English, signing in once with MyGovID, and the portal has to meet S.I. No. 358 of 2020 on web accessibility.
A large corporate dealt with by Revenue’s Large Corporates Division will have to issue EN 16931 e-invoices and report invoice data from 1 November 2028 under the planned timeline, while its ERP still sends PDFs.
A State agency uses an AI system to help staff prioritise grant applications. If it falls under the high-risk rules from December 2027, the agency must be able to show the Irish supervisors, coordinated by the AI Office of Ireland, how it works and who answers for it.
Careers
Architects, engineers, consultants and other specialists who build cloud platforms, enterprise software and AI systems for the public service and for companies around Europe. The problems are real, the users are real, and the people beside you care about getting it right.
Practices
In a first conversation we talk about the problem you want to solve, the systems you have in place and what would help most next. You don’t need to prepare a presentation.
Programmes that cross borders, teams that work in several languages, and one person who answers to you for the whole.
