DataFab //  artificial proprietary intelligence T+00:00:00 ISO/IEC 27001:2022 · SOC 2 Type II — certified Raw egress  0 B 

DataFab  /  Platform  /  Governed utilities

Layer 3 · out of the box

The catalogue, domain by domain.

Ready-made agentic applications that ship on top of the Fabric and the Studio. A utility is agencies, workflows and widgets composed in the Studio and published — the same build surface, packaged for a domain. Each one is grounded in the graph, governed by construction, and enriches the Fabric as it runs.

Grounded in the graphGoverned by constructionEnriches the Fabric as it runs

The pattern

One pattern behind every utility.

A governed utility is not a separate product with its own data model, its own permissions and its own audit story. It is the same platform, scoped to a domain — which is why a second utility costs a fraction of the first, and why every one of them inherits the same guarantees.

01

Grounded

Every agency queries the resolved graph — entities, relationships and provenance — instead of raw, unresolved sources.

02

Composed

Agencies, workflows and widgets assembled in the Studio, on your schemas, then published as a scoped, reusable unit.

03

Gated

The automation dial is set per step: deterministic where the rule is clear, human-gated where judgement belongs.

04

Evidenced

Confidence, trace and governance attached to every decision, captured as the work happens rather than reconstructed later.

05

Compounding

Governed decisions and outcomes flow back into the Fabric — the platform gets better at the work the more work it does.

Fig. 00  //  the reusable pattern behind every domain utility◆ grounded, gated, evidenced
VENDOR CONTROL PLANE reached over PRIVATE LINK (no public path) REGION — [dedicated region · set at deployment] Egress control — allow-list / firewall · default deny AVAILABILITY ZONE A Ingress subnet — private endpoints only[CIDR set at deployment] · no public ingress (P/S) App subnet — data plane computeresolution · engine · gateway · console[CIDR set at deployment] Data subnet — graph storeno egress route · encrypted at rest[CIDR set at deployment] AVAILABILITY ZONE B (HA replica) Ingress subnet — private endpoints only[CIDR set at deployment] App subnet — data plane computehorizontally scaled · same identity model[CIDR set at deployment] Data subnet — graph replicasynchronous / async per RPO[CIDR set at deployment] replicate

The catalogue

What ships ready.

Each utility below runs on the same Fabric and the same Studio, inherits the same governance and audit, and can be extended by your own people rather than by a change request.

01Financial crime & compliance

Alert triage, sanctions and adverse-media screening, and generative compliance — investigation-console ready. Rules express intent over resolved entities rather than brittle joins, so a typology can be written, versioned, replayed against history and tuned before it reaches production.

Open the financial crime utility →

Investigation console — entities, timeline, networkScreening & alert triageCase to filing, narrative linked to sourceGoverned agencies traceable to a regulator
02Due diligence

Counterparty, transaction and onboarding due diligence over the resolved graph. Ownership traversal runs until it reaches natural persons or cannot proceed — and reports the halt, with the level reached and the percentage unresolved, as a finding rather than a silence.

Open the due diligence utility →

Counterparty profile — ownership, control, ultimate beneficial ownerTransaction reviewOnboarding KYC / KYBSourced, defensible report
03Consumer-redress remediation

Remediation at population scale — cohorting affected customers, calculating redress deterministically, running outreach through the AutoFab virtual agent, and packaging the output so it is audit-ready on the day it is produced rather than reconstructed under deadline.

Open the remediation utility →

Cohorting affected populationsDeterministic redress calculationAutoFab outreachAudit-ready packaging
04Client lifecycle management

Client lifecycle and contract management — extraction of parties, terms and dates; obligations tracked and alerted; clause-level review that stays governed; the whole lifecycle from intake to renewal on one resolved graph rather than in a separate contract silo.

Open the client lifecycle utility →

Extraction — parties, terms, datesObligations tracked & alertedClause-level reviewIntake to renewal
05Insurance claim handling

Written claims worked end to end — coverage, prospects, fees and the drafted response, over the systems you already run. Where information is missing it drafts the reach-out; where the rules are clear it resolves; where judgement is needed it prepares the judgement and routes the file to a caseworker with everything attached.

Open the claim-handling utility →

Intake & extraction, per-field confidenceCoverage checking against clauses and endorsementsProspects and statutory fee scheduleDraft & route — a person decides
06National security & intelligence

All-source fusion inside your enclave with zero data egress. Every collection discipline read in place and resolved into one corroborated picture; institutional de-biasing so a warning cannot die inside one analyst’s judgement; horizon scanning against threats defined before they exist; and a governed request lifecycle with tiered authority and deconfliction.

Open the national security utility →

All-source fusion, in placeInstitutional warningHorizon scanningA virtual team per analystAir-gap and tactical edge
07Government intelligence & investigation

Investigation and intelligence for public bodies — grounded and auditable. Entity and network resolution across sources, governed open-source enrichment, case management with chain of custody, and a record produced as the work happens so it survives an inspector, a court or an oversight body.

Open the national security utility →

Open the government investigation utility →

Entity & network resolutionGoverned OSINT enrichmentCase management with chain of custodyFully auditable
08Build your own

Any governed workflow the enterprise needs, composed in the Studio on the Fabric. Compose agencies and widgets, ground them on the graph, govern them with sandboxes, gates and audit, and publish a reusable utility your other teams inherit.

Open the build-your-own utility →

Compose — agencies and widgetsGround — on the Fabric graphGovern — sandbox, gates, auditPublish — reusable utility

Worked example

The Financial Crime Unit, on one living graph.

Financial crime hides in the gaps between systems. Core banking, payments, customer records, screening, trade, documents and dozens of vendor tools each hold a sliver of the truth — and none of them share an entity, a schema, or a story.

01

Fragmented data

The same customer exists a dozen times, reconciled by hand, never by the system.

02

Alert backlogs

Rules fire in isolation; investigators drown in volume, most of it noise.

03

Manual investigation

Analysts stitch the picture across screens, copy-pasting evidence between them.

04

No lineage

When the regulator asks why, the answer lives in someone’s memory.

DataFab — investigation console · case CR-4471grounded · cited · logged
Analyst

Why was customer C-4471 flagged, and who are they connected to?

Investigation agency

C-4471 triggered rapid pass-through: €1.2M received from a high-risk counterparty and moved out within 18 hours across three accounts. Two counterparties resolve to a single beneficial owner already on a watchlist. grounded in: transaction ledger · ownership graph · watchlist — view lineage

Analyst

Show the network and draft the narrative.

Narrative agency

Network rendered as a live widget. Draft narrative assembled with every statement linked to the record it rests on. Filing is held at the gate for your approval. human gate · awaiting analyst authority · the gate itself is logged

Illustrative walkthrough of the utility’s behaviour. Figures shown are example values, not measured results.

Agency

Monitoring

Watch flows and networks for typology hits, continuously rather than in a nightly batch.

Agency

Screening

Resolve names and entities against watchlists and adverse signals — the real entity, not just a string.

Agency

Triage

Rank and dismiss noise with a stated, auditable rationale for each disposition.

Agency

Investigation

Gather evidence across the graph and assemble the case.

Agency

Narrative

Draft the report, grounded in cited evidence, linked back to source.

Agency

Quality

Review, challenge and check the work of the other agencies.

Worked example

The written claim, worked end to end.

Extraction vendors tag the claim and hand it on — and the claim still lands on a caseworker’s desk. Unstructured solicitor letters, coverage judgement and statutory fee-checking stay manual, so the fallout rate barely moves.

DataFab is the layer that finishes the job. It reads the claim and every document behind it, finds what is missing, checks coverage against the policy, assesses the prospects, checks the fees against the statutory schedule, and drafts the grounded response — over the systems you already run, with a caseworker on every decision that matters.

01

Reads & extracts

Per-field extraction confidence, not one score for the whole document.

02

Finds the gaps

Missing information identified and the reach-out drafted; the response re-enters through intake and the case re-evaluates automatically.

03

Checks coverage

Claim facts matched to policy points, exclusions and endorsements — clause by clause.

04

Assesses prospects

Merits and likely outcome grounded in cohorts of similar claims, not a general impression.

05

Checks the fees

Invoices reconciled against the statutory fee schedule.

06

Drafts & routes

A decision-ready file with everything attached — and a person decides.

Fig. 01  //  the fallout pool is where the cost lives● illustrative
Written claimsthe full intake · all life-areasIncumbent extractiontags + extracts · token fee / caseextraction alone does not finish the job35%handled straight-throughroutine · rules-clear · complete65%still falls to a caseworkerat ~40x the extraction fee, fully loaded(peer-validated: 20–75x)the cost is not the extraction — it is the two-thirds that extraction cannot finish
Fig. 02  //  confidence rolls up, never averages away◆ amber marks soft confidence
ONE CONFIDENCE PER DECISION. MANY PER CASE. ROLLED UP, NEVER AVERAGED AWAY.SIGNALS · RAW, PER ARTEFACTOCR field confidence0.96Document completeness0.92Entity match0.95Clause similarity0.83Endorsement scan0.79DECISIONS · ONE PER DECISIONExtraction accepted0.95Policy affiliation0.95Clause match0.81Cohort assignment0.88DETERMINATIONS · CLAIM LEVELCoverage: partial0.74Reserves estimate0.79OUTCOME · CASE ROLLUPWeighted by decision criticality — routed to a personHuman reviewAMBER MARKS SOFT CONFIDENCE — THE CASEWORKER SEES WHERE THE WEIGHT IS THIN BEFORE THEY DECIDE
Trust is the architecture, not a promise

Claims teams have been burned by black-box systems with a blanket accuracy figure printed on the box. This is built the other way up: every output carries its own evidence, so a caseworker can verify any statement in seconds and an auditor can replay any decision end to end. Confidence is calibrated per step, rolled up per claim — one confidence per decision, many per claim.

Domain capability inventory

Twelve domain utilities, not twelve products.

Where a domain is deep enough to warrant it, the utility ships as two layers of domain capability on the shared graph — business utilities that do the domain work, and enterprise utilities that manage the artefacts the domain work produces. Both share the same entities, the same lineage and the same governance rather than reproducing them.

Business utilities — the domain work itself

ScreeningSanctions, politically-exposed persons and adverse signals against the resolved party
Customer due diligenceOnboarding, enhanced diligence, periodic and event-driven review
Adverse mediaGoverned media enrichment, attributed to the entity rather than the string
Client lifecycle managementIntake to renewal, obligations tracked, clause-level review
FraudPattern and inconsistency detection across systems that do not talk
Financial intelligenceMoney-flow analysis, typologies, network and pass-through detection

Enterprise utilities — managing what the work produces

Case managementTimeline, ownership, re-evaluation when new evidence arrives
Alert managementQueueing, deduplication against the resolved entity, disposition with rationale
Rules managementAuthoring over entities, versioning, back-testing, controlled promotion
Document managementDomain documents referenced in place, extracted on request, never duplicated
Source managementWhich approved sources feed which domain field, with credibility and cadence
List managementWatchlists and reference lists versioned, with the list state at decision time

Platform utilities — the building blocks every domain utility is composed from

Entity managementList managementSource managementDocument managementAlert managementCase managementOrchestrationGraph analysisPolicy enforcementAccess managementKnowledge & Agentic StudioReporting engineMCP StudioVision-language-actionComposite AIEntity resolutionGenerative reasoningDifferential privacy

A domain utility is a named composition of these blocks, run by a named agency, over the one resolved graph. Turning one on does not add a data model, a permissions model or an audit trail — it inherits all three.

Build your own

Everyone ships an agent. We ship the studio to build your own.

A generic agent knows generic things. The intelligence that changes your business is yours — built by your people, owned by your firm, running inside your boundary. Encode the expertise once, own it, and deploy it across every team, every client, every jurisdiction.

What it builds

Every run makes the asset worth more.

A utility is not just work performed. Each governed decision, accepted relationship and approved outcome cascades back into the knowledge state with provenance and authority attached. The platform gets better at the work the more work it does — and the holding underneath it compounds rather than depreciates.

Next step

Pick the workflow that hurts most.

A first engagement is one urgent use case, a handful of critical sources, and a governed utility running inside your boundary — with the evidence a regulator would ask for produced as the work happens.