Decision Provenance Framework.

Trace not only where data came from, but how it informed a decision, recommendation or AI response — as one connected chain.

Download the Executive Asset Talk with Data Tiles

Executive summary

Data lineage tells you where data came from. Decision Provenance answers what happened next — which transformations, policies, product version and consumer turned that data into a specific decision or AI output.

In an AI-enabled organization, Decision Provenance is no longer optional. Regulators, customers and executives need a coherent answer to "why did the system decide that?" — and that answer must connect data, policy and outcome.

Why it matters

Without Decision Provenance, AI failures become unsolvable mysteries and AI successes become unrepeatable luck. With it, organizations can debug, defend and improve every decision the system supports.

What leaders should understand

One chain, end to end

Source data → transformations → policies → product version → consumer → decision. Every link is recorded.

It is a byproduct of good design

If governance is active and data products are versioned, Decision Provenance emerges from normal operations.

It is how trust scales

Trust at one decision is anecdotal. Trust expressed as lineage across every decision is a capability.

It is what regulators will ask for

Explainability and auditability requirements are converging on something that looks a lot like Decision Provenance.

Decision Provenance

From data lineage to decision provenance.

Decision Provenance answers what happened: which source data, transformations, policies, product version and consumer produced a specific decision, recommendation or AI response. Decision Integrity answers the next question: whether that decision can be trusted and defended.

Source data→

Transformations→

Policies→

Product version→

Consumer→

Decision

Active Governance

Passive governance documents.
Active governance enforces.

Traditional governance often documents policies, ownership and classifications. Active governance applies those policies at the point where data is created, served and consumed. For AI, this matters because policies governing access, privacy, explainability, provenance and usage cannot sit separately from execution.

Passive governance

• Policies live in documents
• Reviewed quarterly, in arrears
• Enforced by people, when remembered
• Disconnected from execution

Active governance

• Policies live with the data product
• Applied at create, serve and consume
• Enforced by the platform, every time
• Trust is observable, not assumed

How it fits into the Decision-Driven Enterprise

Decision lineage closes the Decision-Driven Enterprise loop. The Workbench produces the data products, Active Governance enforces the rules, and Decision Provenance records what those products did in the world — feeding learning back into the next iteration.

The framework

Decision-Driven Enterprise

The point of view this all sits inside.

The proof

Decision Provenance & Integrity

What happened, and whether it can be trusted.

The language

Glossary

Plain-English definitions for every term.