FoundationAI was founded by Masood Alam and James Nutt, two practitioners who have spent their careers building, selling and delivering enterprise data and AI systems. The thing the LLM revolution has revealed is something the graph community has known for years: software cannot reason about a business it has no model of. We are here to give it one.
An LLM is a remarkably good language predictor. It is a remarkably bad business reasoner. Not because it is stupid, but because nobody has shown it what your business is.
What is a customer at your organisation? Is it a contact in CRM, an account in billing, a party in your ERP, a requester in support? In most enterprises the answer is “all four, slightly differently, with no agreed bridge.” So when the agent answers a question about “customers”, it is making the bridge up. That is the foundation problem.
The fix is not a better model. It is an ontology: an agreed, versioned, machine-readable definition of what the entities in your business are and how they connect. Once that exists, the AI has a map. It can reason. It can cite. It can be wrong in a way you can detect and correct.
If you cannot say what a customer is in a single sentence the whole business agrees with, you do not have a customer agent problem. You have an ontology problem. We start there.
If a CDO cannot explain the ontology to a CFO in five minutes, the ontology is wrong. We model in language the business already uses, then formalise underneath.
RDF, OWL, SHACL, SPARQL, Cypher, JSON-LD, MCP. Every artefact we produce is in a format you could maintain without us. We earn the next engagement by being good, not by being needed.
Your existing data warehouse, lakehouse and operational systems stay. The graph is a layer on top, not a replacement underneath. Most foundations work is connection.
When the AI gives an answer, the path through the graph is visible. The auditor sees it. The regulator sees it. The user sees it. There is no black box.
If a 30 minute call reveals you need a strategy consultancy or a software vendor instead of an ontologist, we say so and recommend who. We are not interested in winning unsuitable engagements.
FoundationAI was founded by two people who have spent their careers in data and AI. Every engagement is led personally by one of them, not delegated to a deck team.
Author of The Data Revolution book series. Winner of the Data & AI Vision Award at CDO Vision 2026. Has built large-scale AI and data ecosystems for government and industry, with a particular focus on semantic architectures and trusted data. Keynote speaker at Big Data London, the CAIO Summit and TechX, and over 10 million content views on data and AI thought leadership.
Sales and delivery leader with experience across large enterprises, mid-sized organisations and startups in the technology sector. Sales leader in energy and utilities, and an expert in AI sales and delivery. Has led large portfolio delivery with teams of 100 plus, with a track record of client-centric, expert-led engagements.
“Data without trust is just noise. FoundationAI exists to give every organisation the foundation to move from noise to confident action, and to build the AI future on solid ground.”
“We are not selling software. We are selling certainty. The confidence for a CDO to say yes to their board, for an organisation to finally deploy the AI they have been promised. That is the real product.”
“The foundation work was the bit nobody had wanted to do. Six months in, every customer answer the AI gave traced back to a node in the graph. Our auditors stopped asking how we knew.”
We will not pitch. We will ask hard questions about your ontology and tell you honestly whether we are the right people to help.
Book a meeting →