Agentic analytics, engineered for trust

Ask your data anything. Trust every answer.

Data Foundations lets anyone ask questions in plain language and get governed, verifiable answers, because the AI reasons over your business concepts, never raw tables. No hallucinated metrics. No guesswork. Just clarity your whole organisation can act on.

Book a demo Read our story
You ask
"Revenue by region this year?"
Plain language
→
Mapped to concepts
Revenue · Region · Year
Governed ontology
→
Trusted answer
Verified · auditable
Compiled to SQL
0
Questions answered
0
Answers traced to source
0
Warehouses supported
0
Faster time to insight
Our story

We built a foundation for AI you can actually trust.

Most AI analytics tools let a language model write SQL directly. The result is unpredictable queries, hallucinated metrics, and no guarantee that "revenue" means the same thing twice.

We started from a different premise: an analytics system that speaks your language, understands your domain, and gives answers you can stake a decision on. At the heart of Data Foundations is the Ontology Graph, a semantic layer that captures your metrics, governance rules and definitions.

The AI only ever reasons over those governed concepts. Every question is translated into DFQL, our query language, and compiled deterministically to SQL with your business logic applied. Even the chart titles and labels are rendered from DFQL, not invented by a model. No hallucinations, just clarity.

Data Foundations · live answer
What's revenue by region this year?
DData Foundationsverified · DFQL
£4.2MEMEA
£3.6MAPAC
£3.3MAmericas
£2.3MUK
£1.6MAPAC-S
What makes us different

Three principles behind every answer.

The same ideas that guide the product guide the company: be trustworthy, be transparent, and put understanding in everyone's hands.

01

Governed by an ontology

Your metrics, relationships and rules live in the Ontology Graph, a semantic layer the whole organisation shares. Definitions stay consistent, so "revenue" is the same number every time.

02

No hallucinated metrics

The AI writes DFQL against governed concepts, never raw SQL. A deterministic compiler turns it into SQL with your business logic applied, so results are exact and auditable.

03

Secure by design

Because the AI only sees concepts, it can't be jailbroken into reaching raw tables or unauthorised data. Row and column security travel with every question.

What we value

Curiosity, clarity, and answers you can defend.

Curiosity for everyone

The best insights start with an offhand question in a meeting. We build for the spark of curiosity, not just the trained analyst, so anyone can explore without waiting on a dashboard.

Trust as a feature

An answer is only useful if you can act on it. We treat transparency, lineage and governance as core product, not an afterthought, so every number can be traced back to its source.

Clarity over cleverness

We'd rather be understood than impressive. Plain language in, plain explanations out, with the working shown. The product should feel like magic and read like a sentence.

The problem

Data locked behind specialists

Business teams waited days for answers while questions piled up. The tools that promised self-service still demanded technical skill.

The idea

What if you could just ask?

A system that speaks your language and understands your domain, giving answers you can trust, regardless of your technical background.

Today

Agentic analytics, in production

Teams ask millions of questions and get governed answers inside the tools they already use, from a chat panel to the API.

Ready to make your data answerable?

See how Data Foundations turns plain-language questions into governed, verifiable answers your whole team can trust.

Book a demo Back to top