Your AI has your data. Now give it your meaning.
Folaint builds the your AI checks before it answers: grounded in your industry’s open standards, not guessed from your data. Your experts approve every fact. You own the result.
See it in your industry
Pick an industry. The pattern is the same every time: a question that costs days today is answered in seconds, and every part of the answer links to the document it came from.
Illustrative examples. Names, figures and documents are fictionalised.
Every common format, connected into your knowledge graph.
PowerPoint, Excel, PDFs, SAP and the tools you already run on: Folaint reads the formats and systems your teams live in and turns them into one governed knowledge graph. New connectors land with every release.
All product names and trademarks are the property of their respective owners. Names are used to describe data-source compatibility and do not imply any affiliation or endorsement.
Three of the questions above, tried against the usual fixes.
The two common fixes are a context file (instructions in the prompt) and RAG, top-k nearest-neighbour retrieval over vector embeddings. Both help. Neither survives the three questions you just saw.
“Which delivered machines contain valve V-220?”
Vector search ranks. It doesn’t enumerate.
The usual fixTop-k nearest-neighbour search returns the 10 to 20 most similar passages: it maximises relevance, not completeness.
What it needsA set-membership query: traverse the bill-of-materials graph and return all 47 machines, with proof there’s no 48th.
For a product recall, “probably most” is a wrong answer, and no tuning of k changes that.
“How much exposure do we really have to Mr. K.?”
Retrieval is extractive. This is a computation.
The usual fixThe €3.1M figure is written in no passage. Extractive retrieval returns only text that already exists.
What it needsResolve “Mr. K.” to one entity across three systems, then aggregate the exposure: entity resolution plus a graph join.
An embedding can only return what was written down somewhere, and this number never was.
“Finance says 14%, Sales says 19%. Which margin is right?”
A prompt is context, not a constraint.
The usual fix“Net of rebates” is just tokens in the context window; the model can weight the Sales table higher and answer 19%.
What it needs“Margin” resolves to one typed metric: a canonical formula with an owner, a version and rules every answer is validated against.
Instructions the model may ignore aren’t governance, and no one keeps a thousand of them current in a file.
Context files and RAG give your model more text. Folaint gives it verified facts, complete relationships and enforced definitions it can query.
Your AI can’t reason over rows. It can reason over a map of meaning.
A gives your AI entities, relationships and rules it can follow, anchored in your industry’s published standards, not just patterns inferred from past data.
Live graph: drag any node, hover a dashed one to see what stands behind it.
Five steps, your people in the loop at every one.
The video walks through the story end to end. Below is the machinery inside each step. One thing holds throughout: an automated pipeline does the building, and your experts appear exactly three times, at the approval gates, to check what it proposes rather than model anything themselves.
- 01
Start
Ontology design starts from your industry's open standard: for finance, for construction, for manufacturing. The pipeline loads and maps it automatically: entity types, properties and permitted relationships, thousands of terms your industry has already agreed on. That draft becomes your ontology, the formal vocabulary the whole graph is built on. Nobody on your side models anything; your experts will only correct it later.
Under the hood- The standard is loaded as the ontology skeleton (TBox): classes, properties and the relationships permitted between them.
- The standard is the prior, not your documents: it carries distinctions regulation requires that your data never wrote down.
- It loads from a local bSDD/OWL export or the bSDD API, mapped straight into the graph schema. No modelling by hand.
DeliversA draft ontology for your sector, anchored in a published standard, before a single internal meeting. - 02
Surface
Nobody fills in templates and nobody sits in workshop marathons. The pipeline scans your documents and systems and proposes two things. First the competency questions: the concrete questions the graph must be able to answer, like “which contracts renew next quarter, and on what terms?”, each stored with its expected answer as a permanent test case. Second the definitions, exceptions and rules it finds behind them, each with its source attached. Your experts review a finished list in one sitting. This is where knowledge graph projects usually stall for months; here it is the fast part, because the drafting is machine work.
Under the hood- Each competency question is stored with its expected answer and a test query the finished graph has to satisfy.
- Two elicitation paths, one output: a guided expert interview and mining your documents both produce the same question set.
- Definitions and exceptions are surfaced from your data, each carrying the source document and page it came from.
DeliversA set of competency questions with expected answers, plus every definition and exception: proposed by the pipeline, reviewed by your experts. - 03
Confirm
Before any data flows, the draft ontology is tested against every competency question: can it answer each one, in the right shape? Then comes the human gate, and this is the whole of what we ask of your experts: review what the pipeline proposed, correct it where the business disagrees, approve. Nothing enters the until a named expert has signed it off. This gate is deliberate: a wrong definition that looks official does more damage than a missing one. Every element of the ontology exists to answer a real question, and every fact traces back to the person who approved it.
Under the hood- Ontology evaluation: every competency question runs against the schema. Can it answer each one, in the right shape?
- SHACL rules act as a deterministic gate: anything malformed is caught by machine, so a human only ever sees valid candidates.
- The human gate is one of exactly three: a named expert reviews, corrects and signs off before anything counts as true.
- Each decision is stored with its owner and timestamp, so the authoritative definition can always be traced to a person.
DeliversAn evaluated, expert-approved ontology. Every element serves a question; every fact has a named owner. - 04
Connect
Now the pipeline fills the graph, unattended. Connectors read your documents and systems (PDF, Word, Excel, SAP exports, IFC models), and extraction runs in two passes: first the , then the relationships between them, stored as triples. Every triple carries its source document, page and timestamp. Duplicates are resolved: the same customer under three names in three systems becomes one record with a stable identifier. Nobody on your side keys anything in; at the second gate your experts review a sample of the extracted facts, one by one, before the full load runs.
Under the hood- A connector per format: PDF, Word, PowerPoint, Excel, email, IFC/BIM models, SAP exports. Scanned pages go through OCR.
- Two-pass extraction: entities first, then the relationships between them, written as triples (subject, relation, object).
- Entity resolution: three names in three systems collapse to one stable identifier, so your AI stops counting them as three.
- Provenance on every triple: source document, page, timestamp and a confidence score. This is what makes each answer traceable.
- A sample-validation gate before the full load: your experts review a batch of extracted facts, one by one.
DeliversA populated knowledge graph: every fact traceable to document and page, every real-world thing represented exactly once. - 05
Keep true
A graph no one maintains drifts away from the business it describes. So maintenance is a pipeline, not a promise. When a source document changes, it deletes exactly the facts derived from it and re-extracts that one document, nothing else; the provenance on every triple is what makes that targeted update possible. After every change to the ontology, all competency questions run again as regression tests, automatically. And every version is kept, so you can see what changed, and when.
Under the hood- Delta re-ingestion: change one document and only the facts derived from it are deleted and re-extracted. The provenance makes this surgical.
- CQ regression: after any change to the ontology, every stored competency question runs again, automatically, as a test.
- Versioning: the ontology carries a version tag and every version is kept, so you can see what changed and when.
- Temporal confidence: facts carry a valid-from and valid-to, and trust decays over time, so stale meaning is flagged, not trusted.
DeliversA layer that stays correct as your business changes: targeted updates, automatic re-testing, full version history.
What the words actually mean.
Standards-anchored, not precedent-based.
It captures distinctions that never appeared in your data but that regulation requires. An and rules make the meaning explicit and checkable.
Open, not a private dialect.
Built on standards you could read, export and keep, never a black box only we can interpret.
- Knowledge graph
- A network of entities (customers, contracts, orders) connected by typed, named relationships. Unlike database rows, every connection carries a defined meaning an AI can follow and reason over, not just a foreign-key join.
- Ontology
- A formal vocabulary for a domain: the entity types, their properties and the relationships permitted between them. Think of it as a schema for meaning, not just for data: “a Contract must have a counterparty of type Customer.”
- Triple
- The atomic fact in a knowledge graph: subject, relationship, object. “Contract 2024-117, has counterparty, Acme GmbH.” Extraction from your documents runs in two passes, entities first, then the relationships between them, and every triple stores the source document, page and timestamp it came from.
- Competency question
- A concrete question the graph must be able to answer, collected with your experts before anything is built: “which suppliers are single-source for a critical part?”. Each one is stored with a test query. After every change to the ontology, all of them run again, a regression test for meaning.
- SHACL
- Shapes Constraint Language: a W3C standard for writing rules that data in the graph must satisfy. Machine-readable and machine-checkable: if a compliance rule says every supplier must carry a risk rating, SHACL will flag any that don't.
- Open standards
- The W3C publishes the full specifications for RDF, OWL and SHACL. Any developer can read, audit and implement them independently. Your data model is never proprietary. You can export it, inspect it, and take it elsewhere without our involvement.
Two assets you own. Every door they open.
At the end of the process you hold a that maps your domain and an that governs it. Both in open formats, running on your infrastructure. Below are the capabilities they hand you, regardless of which AI layer you choose next.
Entities, relationships, and connections your AI can traverse and reason over.
The definitions, rules, and governance layer that makes every fact in the graph trustworthy.
Power any AI stack
Your LLM, pipeline, and agents all draw from the same governed graph. Connect new tools as you adopt them. The meaning you built is reusable across every layer.
Audit and certify every output
Every graph-derived answer carries a reasoning path back to the source document and the expert who confirmed it. Compliance reviews and regulatory audits become a lookup, not an investigation.
One record across every system
Entity resolution gives every customer, product, and asset a single canonical ID. Your CRM, ERP, and support desk stop disagreeing. Master data governance becomes a side effect of the graph.
Swap models, keep your work
The captures what your domain means. Change LLM providers, upgrade versions, or add new reasoning engines. Your investment in the graph survives every model generation.
Open formats, full portability
The graph and are stored as standard RDF/OWL. Export them, host them on your own infrastructure, hand them to a new partner, or migrate them at any time. No proprietary format holds you.
A compounding knowledge asset
Every expert review sharpens the . Every resolved entity makes the graph denser. Unlike statistical model weights, this asset encodes deliberate choices and accumulates value as your domain evolves.
Most companies build on rented intelligence: when the model changes, meaning disappears. Your Knowledge Graph and Ontology stay. Both yours, in open formats. Rent the intelligence anywhere. The meaning you own, with us.