Data Governance & Quality
Most governance programmes produce a catalogue that is accurate on the day it is published and decays from there, because maintaining it is somebody's eleventh priority.
We build governance that is largely automatic: metadata harvested from the platform, lineage derived from actual query history, quality rules executed in the pipeline, and stewardship assigned to people who already own the underlying process.
Platforms we typically deliver this on
Scope of the service
Governance framework
Policies, roles, decision rights and the forum that resolves disputes — sized to your organisation rather than to a textbook.
Data catalogue and metadata
Automated harvesting with business glossary, owners and classification, kept current by the platform rather than by hand.
Lineage
End-to-end lineage from source system to dashboard, derived from query history and pipeline definitions.
Data quality management
Rules defined with business owners, executed in-pipeline, with scorecards and remediation workflow.
Classification and privacy
Identifying and tagging personal, sensitive and regulated data, and enforcing handling rules including residency and retention.
Master and reference data
Golden-record design and survivorship rules for the entities that matter — citizen, patient, student, customer, vendor.
Where this is used
Representative engagements, described at the level our clients permit. Sector and shape are accurate; identifying detail is withheld.
Classifying personal health information across the estate
The organisation could not state with confidence where PHI existed. Automated scanning and classification across databases and file stores produced an inventory, then drove access and retention rules.
Outcome — A defensible inventory of where regulated data actually lives.
One definition of an enrolled student
Six departments counted enrolment differently, so no institutional number was trusted. We facilitated agreement on definitions and implemented them once in the semantic layer.
Outcome — A single enrolment figure every department can reconcile to.
Lineage for audit and information requests
Auditors asked where a published figure came from and the answer took weeks. Automated lineage now answers it from the catalogue.
Outcome — Provenance questions answered in minutes with evidence.
Before you get in touch
Which catalogue tools do you use?
Unity Catalog, Microsoft Purview, Collibra and open-source options, chosen against your platform. We favour tooling that harvests automatically over anything requiring manual upkeep.
How do you get people to maintain the catalogue?
Mostly by not requiring them to. What must be human-supplied is limited to business definitions and ownership, and those are collected once during onboarding of each domain.
Is governance required before AI?
Not all of it, but classification and access control are. You cannot safely put AI over data when you cannot say what is in it or who may see it.
Often engaged alongside this
AI Governance & Responsible AI
Frameworks, controls and evidence built for organisations that have to explain their AI systems to a regulator, an auditor, a board or the public.
Read moreData & AnalyticsData Strategy & Consulting
Target-state architecture, operating model and a sequenced modernisation plan — including the systems and reports to retire.
Read moreData & AnalyticsData Engineering & Pipelines
Ingestion, transformation and orchestration built with tests, lineage and alerting — so a broken feed is an alert, not a discovery three weeks later.
Read moreStart with three weeks and a straight answer
The AI Readiness Assessment is fixed in scope, fixed in price and produces four deliverables you own — whether or not you continue with us.