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

DataFab  /  Platform

One platform · three layers

Inside the platform, layer by layer.

DataFab is a foundation, a builder and a set of governed utilities on one platform. The foundation draws the enterprise from the systems already running. The builder lets the enterprise formalise and execute how it works. The utilities ship ready. And every governed result flows back, so the whole thing compounds.

Read in placeSchema-boundedHuman-gatedTamper-evident audit
Fig. 00  //  the sovereign data foundation● connect, don’t copy · raw egress 0 B
DataFab Sovereign data foundation One resolved truth, inside your boundary CONNECT, DON’T COPY RESOLVE IN PLACE · NO RAW EGRESS ENTITLEMENT REFUSED · NOT TRAVERSABLE One resolved enterprise SOURCED · GOVERNED · REPRODUCIBLE EVERY TRAVERSAL LOGGED COMPOSE ↓↑ OBSERVE & REPLAY Knowledge & Agentic StudioPLAIN LANGUAGE · NO CODESCREEN · MONITOR · INVESTIGATE · ASSUREBUILT BY DOMAIN EXPERTS GATE · PROPOSE, DISPOSE Resolved entitiesGoverned accessLineage & replayContinuous monitoringCase investigation READS · GOVERNED WRITE-BACK ① NOTHING ENTERS UNENTITLED   ② NOTHING ACTS UNGATED CORE ACLCRM ACLKYC ACLREGISTRY ACLPAYMENTS ACLCONSOLIDATED MODELINHERITED ACLs ARE THE FLOORSTRICTER, NEVER LOOSER

Everything, working together

The whole machine.

Sources stay in place and are discovered and resolved into one governed knowledge state. Agencies, built in the Studio, ground on it and run — with people at the gates. Utilities ship on top. And every governed result flows back, so the knowledge state is cumulative, not static.

Step one

Connect the enterprise

Point the fabric at a source you already run. Direct database, API, event stream or MCP. No extract, no copy, no migration, and no system asked to change.

Step two

Understand it

Discovery catalogues the estate. Entity resolution collapses the twelve copies of one customer into one golden record. A schema-bounded ontology is derived from what was found.

Step three

Build how it works

Your own people compose governed agencies in the Studio — in language or with full control — and publish them, scoped, sandboxed and audited.

Step four

Compound every result

Governed schemas, accepted relationships, approved decisions and outcomes cascade back into the knowledge state with provenance, authority and version attached.

Fig. 01  //  the whole machine — and every governed result flowing back
LAYER 3 · UTILITIESLAYER 2 · AGENTIC STUDIOLAYER 1 · KNOWLEDGE FABRICSOURCE SYSTEMS · IN PLACECRMCase MgmtERPDocumentsEmailOSINTresolve in place → one governed knowledge stateAgencyAgencyAgencyHuman gateapprovalanalystFinancial CrimeDue DiligenceRemediationLegal CLMGov Intel+ yoursDISCOVER · RESOLVEGROUNDENRICH · GOVERNEDCOMPOSE · RUNEVERY RESULT ENRICHES
Fig. 02  //  the reusable pattern behind every domain utility
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

Layer 1 · the foundation

Knowledge Fabric

A data fabric that gives unified access to distributed data assets — maintaining mappings and relationships rather than duplicating data. It is an access enabler, not a repository: one source of truth, unified analytics, nothing moved.

Discovery

Crawlers and profilers catalogue the estate — metadata and statistics only, no manual inventory.

Entity resolution

Blocked, scored, clustered, reviewed and merged into one authoritative golden record with full provenance.

Derived ontology

The semantic layer emerges from your sources — schema-bounded, human-approved, versioned.

Persistent graph

Corporate memory as a four-level knowledge tree: attributes, relations, keywords, communities.

Active metadata

Classification, quality monitoring, lineage and change detection — the knowledge state maintains itself.

Governed enrichment

A registry of approved external sources, each field schema-bound and traced from source to graph.

Layer 2 · the builder

Knowledge & Agentic Studio

The build surface of DataFab — where the enterprise composes its own governed agencies from its own documents, in its own language, and on its own schemas. Every agent schema-bound, sandboxed, human-gated and audited.

Business user · no code

Build it in language.

  • Describe it in words — planning drafts the agent
  • Upload your documents — discovery extracts the schemas
  • A sentence becomes a pipeline you can preview
  • Start from a template of pre-built workflows

Expert · full control

Compose it node by node.

  • Visual builder with live validation on every node
  • Custom logic and scripts inside the query plan
  • Author your own connectors in MCP Studio
  • Define schemas and query plans directly
  • Step-by-step debugging with breakpoints and traces

Both routes converge on one governed agent. A pipeline generated from words imports straight into the visual builder, and the same draft-and-publish lifecycle carries it between them.

Layer 3 · out of the box

A utility is the platform, pointed at a problem.

A utility is agencies, workflows and widgets composed in the Studio and published — the same build surface, packaged for a domain. As it runs, its governed decisions and outcomes enrich the Fabric, so the platform gets better at the work the more work it does.

Available

Financial crime & compliance

Alert triage, sanctions and adverse-media screening, investigation console, and case to filing with the narrative linked to source.

Available

Due diligence

Counterparty, transaction and onboarding due diligence over the resolved graph — ownership, control and ultimate beneficial owner.

Available

Car-finance remediation

Cohorting affected populations, deterministic redress calculation, AutoFab outreach and audit-ready packaging.

Available

Client lifecycle management

Extraction of parties, terms and dates; obligations tracked and alerted; clause-level review; intake through renewal.

Available

Insurance claim handling

Intake and extraction with per-field confidence, coverage checking against clauses and endorsements, prospects, fees, drafted response.

Available

Government intelligence

Entity and network resolution across sources, governed OSINT enrichment, case management with chain of custody.

How much it runs itself

The level of automation is a dial.

DataFab does not force one mode. Each use case — and each step within it — runs at the level of autonomy its risk allows. You put the automation where the work is routine, and keep the control where the risk is.

Test drive  //  the automation dial● drag it, or arrow-key it
Level 01 of 05

Deterministic

Fixed steps, reproducible, no model discretion in the flow.

Where it fits
Fee checks, redress calculations, statutory schedules, fixed procedures.
Model discretion
None. The rule decides, every time, identically.
What it leaves behind
Reproducible to the penny. Replayable under the rule version in force.
Level 02 of 05

Process-governed

A defined business process routes the work; rules drive the flow.

Where it fits
Regulated workflows, service-level agreements, sign-off chains.
Model discretion
Routing only, inside a declared process.
What it leaves behind
The process is the audit trail. Every branch taken is recorded.
Level 03 of 05

Human in the middle

Agencies do the work; a person approves at each gate.

Where it fits
Coverage decisions, conflicts, filings, anything contestable.
Model discretion
Prepare and propose. Never dispose.
What it leaves behind
The gate blocks until approved, and the approver and their basis are logged.
Level 04 of 05

Self-organising

Agencies route and adapt within policy and schema bounds.

Where it fits
Triage, enrichment, discovery, evidence gathering.
Model discretion
Chooses its own path — inside the bounds you set.
What it leaves behind
Every path taken is recorded; the bounds themselves are versioned.
Level 05 of 05

Fully agentic

Autonomous reasoning inside the sandbox and the schema.

Where it fits
Extraction, summarisation, low-stakes work.
Model discretion
Full, within an isolated runtime and a bound schema.
What it leaves behind
Post-hoc audit. It cannot invent entities outside the schema.

And it is set per step — one written claim, eight decisions, five different levels

Read the documents
Extraction runs agentic inside the schema — it flags rather than invents.
Find what is missing
Self-organising: it decides which records the case still needs.
Check coverage
Human in the middle. A caseworker approves the coverage call.
Assess prospects
Human in the middle. The agency prepares the judgement; a person makes it.
Check the fees
Deterministic against the statutory schedule. No discretion.
Calculate redress
Deterministic. Reproducible to the penny.
Draft the response
Self-organising within the bound schema, held at the gate.
Issue the decision
Human in the middle. Always.

Connectivity · MCP First

Every system, one protocol layer.

The fabric both consumes and generates Model Context Protocol connectors for federated queries. When a firm-specific system has no built-in connector, you do not wait for the platform — you author one, have it validated, and publish it to your own catalogue.

PostgreSQLMySQLMongoDBNeo4jCloud data warehousesLakehouse tablesClickHouseOracleSQL ServerSalesforceHubSpotSlackGmailGoogle DriveJiraGitHubAirtableNotionAWS S3Azure BlobGoogle Cloud StorageDropboxOneDriveSharePointiManageNetDocumentsBoxAderantEliteClioPracticePantherChromaDBLanceDBEvent streams
Connection patternUse caseSeverityRemediation SLA
Direct TLSCloud-hosted sourcesCritical — flow stopped24 hours
VPN tunnelOn-premises sourcesHigh — degradation72 hours
Private linkSame-cloud sourcesMedium — workaround exists7 days
Agent-basedAir-gapped environmentsLow — cosmeticNext release

Release monitoring is continuous, breaking-change alerts are issued as announced, compatibility is reviewed per vendor release, and pre-release testing is performed where the vendor makes it available.

FabriCloud · the private data cloud

Bring the cloud to your data.

FabriCloud is the infrastructure layer DataFab runs on, and a product in its own right: cloud-grade data and AI — analytics, applications and agents — running inside your own environment. Not a public cloud. Not a data warehouse. A private cloud that comes to your data, bringing cloud-grade services to the data you already control.

100%Yours — keys and control
0Raw-data bytes leaving your environment
1Governed layer across every source
Apps and agencies on top

The common approach

Many sources × many use cases = endless projects.

  • Every source wired to every jurisdiction wired to every use case
  • Each line a separate project
  • Each re-solving data quality, governance, context and cost
  • It never ends

FabriCloud

Woven once · governed once · served everywhere.

  • Every source — databases, APIs, documents, streams — woven into one fabric
  • Only metadata and control signals leave; raw data is processed in place
  • A data plane stands up in hours; first sources typically same-day
  • Cloud, VPC, on-premises or fully air-gapped, control plane self-hosted where required
Analytics

Trusted by default

One governed, lineage-tracked view for business intelligence and reporting — without another migration.

AI agencies

Permissioned context

Agencies act only on data they are allowed to see — governed tools, full audit. MCP First: every governed dataset, connector and tool exposed as a native Model Context Protocol tool, with access enforced at the source.

Applications

Ship features, not plumbing

Real-time, access-controlled data through one uniform API.

How the two fit together

FabriCloud is the private data cloud — the boundary, the planes, the governed activation layer. DataFab is what runs on it: the Knowledge Fabric that resolves your estate, the Knowledge & Agentic Studio your people build in, and the governed utilities on top. Same architecture, same certifications, same boundary raw data never crosses.

Deployment

Deploy it where you need it.

The same governed platform, delivered against your estate, your controls and your operating model. The control plane is an operating model, not a place your data goes; the data plane runs inside your environment.

Cloud

Single-tenant, your region

DataFab-hosted, single-tenant, deployed in the region you choose, under your residency rules.

Private cloud

Your account, private link

Runs in your own cloud account and VPC, over private connectivity, with your keys.

On-premises

Inside your data centre

Deployed on your infrastructure, reading in place to the systems you already operate.

Air-gapped

Isolated and sovereign

Full functionality inside an isolated network. Agent-based and outbound-only where a boundary must be crossed at all.

Managed service & BPO

Price the outcome, not the headcount.

Delivered as a managed service or inside a business-process operation, the governed agencies do the work and a person sits on the gate. That moves pricing from per full-time equivalent to per case — and because the fabric compounds, unit cost falls as volume grows instead of staying flat.

Today · priced per person

100% of cost per case

  • Cost driven by analyst time
  • Margin competed away head by head
  • Unit cost flat as volume grows
  • More throughput means more hires

Governed · priced per case

A designed margin, with a floor

  • Agencies do the work; the analyst sits on the gate
  • Margin designed in rather than competed away
  • Unit cost falls at volume as the fabric compounds
  • Throughput scales without headcount

Illustrative model, not a quote. The shape of the economics is set per engagement against your case mix, volumes and clearance policy.

What it builds

Three layers. One asset.

Every layer of this platform exists to build one thing: resolved institutional context that belongs to you. The foundation makes it reachable, the builder makes it executable, the utilities put it to work — and each governed result makes the asset worth more than it was the day before. That is the difference between a tool you renew and an asset you hold.

Next step

See it running against your own estate.

One urgent use case, a handful of critical sources, and a governed agency live inside your boundary. That is the shape of a first engagement.