Open operating framework · community edition

Your data already exists. What's missing is who answers for it.

Numbers that don't match across departments, nobody knows who approves access, the audit asks for proof and there are only emails. KDV turns data governance into routine: every shared dataset gets an accountable owner, an agreement of use, rule-based access and evidence of the value it creates.

An open method. Apply it with your existing tools or choose an implementation suited to your context.

Puts existing practice to work DAMA-DMBOK 2 DCAM Data Mesh

Do you recognise yourself?

Choose your starting point. The answer appears right here.

By your role or by your type of organisation. If you like, save your profile: the Documentation will then suggest a reading path for you.

Business area manager
“Every meeting starts with an argument about which number is right.”
What changes
A data product with an owner and an agreed definition replaces parallel versions of the same number.
Where to start
Choose an important decision and sponsor the first case.
Suggested reading
See the Playbook

Where do I start?

Three questions to suggest a profile and a reading path. Nothing is saved.

Data team size
Maturity
Main goal

Small team and starting out: Solo. Several areas or an established process: Organization. Otherwise: Team. Your goal selects the guides.

Suggested profile: Solo

Start with a small scope and clear responsibilities.

Applying the framework does not require buying the product. KDV product: an optional implementation

What KDV does

Data governance as everyday work, not a separate project.

Start with a single use case. Every delivery records who decided, what was agreed and what value it created.

  1. 01 · Owner

    Someone answers for every shared dataset.

    An owner and a steward named for each business domain and data product.

  2. 02 · Agreement and rule-based access

    One agreed definition and time-limited access.

    A versioned contract states what the data means and what it may be used for; the platform enforces the access policy.

  3. 03 · Evidence

    Proof produced by the work itself.

    Decisions, tests and usage leave evidence for review. They support assessment and improvement without producing an automatic maturity score.

How it works

Four stages, from raw data to observed value.

Choose a stage to see what to do, who takes part and what evidence is kept. Before them, a short preparation scopes the first case.

Cultivate

Prepare with purpose.

Starts when

A defined use case and an identified source.

What to do

  • Connect the sources
  • Document the origin
  • Prepare the data
  • Establish a reproducible delivery

Evidence kept

  • Source inventory
  • Record of transformations (lineage)
  • Initial quality tests

Who

The minimum every shared dataset needs

Five common items. Everything else, each organisation decides.

Role names, approval flows and review rhythm can be adapted. These five items cannot be removed.

Role namesApproval flowsReview rhythm
01Named owner

Without an owner, nobody answers for errors, changes or access requests.

How to check

Every shared product has a named owner and steward, with the role recorded.

Evidence

Domain charter and role register.

Template

Domain charter
02Versioned contract

Numbers diverge when each area uses its own definition.

How to check

There is a current contract with version, definitions, quality and permitted use.

Evidence

Published contract and version history.

Template

Data product contract
03Classification

The rule of use depends on how sensitive the data is.

How to check

Each product has a sensitivity level (public, internal, confidential, restricted).

Evidence

Classification in the contract and in the catalogue.

Template

Data product contract
04Policy-based access

Access granted by email is rarely revoked.

How to check

Access only goes through if the policy allows it; every grant has an expiry and a review.

Evidence

Access decision, applied policy and expiry date.

Template

Access review list
05Auditability

Audits and data protection law ask for proof, not intent.

How to check

You can reconstruct who decided, what was applied and what was used.

Evidence

Log of decisions, access and usage.

Template

Evidence register

Why a hive?

We use your terms first. The hive only helps you remember.

Hives cooperate without central command: each looks after its own and respects common rules. It is an organisational metaphor; it does not replace your organisation's terms or assign roles based on bee biology.

Select a row to see the concept without leaving this page.

What your organisation calls it: EcosystemKDV term: Ecosystem

The boundary of collaboration: who takes part, for what purpose and under which common rules.

Example: A network of regional councils sharing professional registration data; a business group with several companies; a municipality whose departments exchange indicators.

Why this name: The natural environment where many hives live side by side and depend on one another.

What your organisation calls it: Business domainKDV term: Hive

The business area accountable for the meaning and priorities of a set of data.

Example: Finance; Customer service; Professional registration; Inspection.

Why this name: The hive: an autonomous unit with its own organisation that contributes to the whole.

What your organisation calls it: Business subdomainKDV term: Cell

The sub-area, team or process where data is prepared and from which products come.

Example: Accounts payable (in Finance); Enforcement notices (in Inspection).

Why this name: The cell: where the work happens and the product is stored.

What your organisation calls it: Business groupingKDV term: Comb

An optional grouping of subdomains within the same domain, for shared rules or permissions.

Example: “Revenue”, bringing together Billing and Invoicing within Finance.

Why this name: The comb: a set of cells forming a structure.

What your organisation calls it: Data productKDV term: Data product

A trustworthy set of data, published with an owner, a contract and known consumers.

Example: Weekly order aggregate; professional register API; inspection prioritisation dashboard.

Why this name: The honey: a useful result, labelled with origin and validity (the contract).

What your organisation calls it: Execution environmentKDV term: Farm

Where products are prepared and operated: production, staging, development.

Example: Production; Staging; Development.

Why this name: The apiary: where hives are placed and tended by the beekeeper.

What your organisation calls it: Prepare · regulate · publish · use and measureKDV term: Cultivate · Govern · Share · Activate

Example: The four stages of the Playbook.

Why this name: The beekeeping cycle of care.

Open method, optional implementations

An open method. Choose the support that fits your context.

The Framework content is available under CC BY 4.0. Apply it with agreed working practices and technical capabilities suited to your context. The method does not require a specific platform, supplier or commercial service.

01

Start with a small team

Documents, a controlled spreadsheet and a review meeting establish owners, agreements and decisions.

Start with the Playbook
02

Use the platform you already have

The method connects to the existing catalogue, identity provider and data platform.

Start with the Playbook
03

Combine modular capabilities

Combine the capabilities for storage, data preparation, catalogue, access, consumption and operations that the case requires.

Start with the Playbook

Illustrative examples

Different contexts. The same questions.

Who is accountable? What was agreed? Who may use it? What value did it create?

From conflicting sales reports to an agreed revenue product.

Finance and Sales agree definitions, refresh, an owner and a versioned contract. The board discusses the decision, not the number.

  1. CultivateInventory of invoicing and CRM sources
  2. GovernA single definition of “recognised revenue” in the contract
  3. SharePolicy-based access for commercial teams
  4. ActivateIndicator: meetings without manual reconciliation of numbers

Fictitious scenarios. They are not customer cases or delivery commitments.

Next step

Pick a hard decision and make it your first data product.

The Playbook guides the preparation and the four stages, with templates ready to download.

navigateEnter openEsc close