Active experiments running

Governance that's tested, not theorized.

The Problem
Most governance fails before it reaches production — because it was never stress-tested in real conditions.
The Principle
Build it in the sandbox first. Understand the failure modes. Then bring it to the organization.
The Work
Four active sandboxes — AI, Analytics, Architecture, and General Governance — each a live experiment.

This is not a consulting site. It's a working lab — a place where data governance ideas get built, broken, and rebuilt until they're actually useful.

Enter the sandboxes
// the problem

"Governance frameworks are written for the ideal state. The sandbox is where you meet the actual state."

Most organizations don't have a governance knowledge problem. They have a governance application problem. Frameworks get adopted before they're understood, and abandoned before they're useful. The lab exists to close that gap — deliberately, methodically, in public.

// active sandboxes

Four domains.
All under active testing.

Each sandbox is a structured experiment — a real problem, a working approach, and honest findings. Nothing here is finished. That's the point.

01 / 04

AI Governance Sandbox

Test AI governance before prescribing it. Who is accountable when the data source is a model? What breaks when AI outputs enter a governed pipeline without a defined owner?

Live
02 / 04

Analytics Sandbox

Organizations trust dashboards they can't explain. The principle here is lineage before launch — no metric goes to a decision-maker without a tested definition, a known owner, and a visible chain of custody.

Active
03 / 04

Data Architecture Sandbox

Architecture decisions are governance decisions — most organizations just don't realize it until migration. This sandbox stress-tests how structural choices either enable governance or silently undermine it.

Building
04 / 04

General Data Governance Sandbox

Governance problems that resist clean categorization. Ownership disputes, policy design that nobody follows, stewardship without authority — the hard organizational friction that frameworks never account for.

Active

Donald Williams

// Founder · Data Governance Lab
Approach
Diagnose first. Prescribe second. Never the other way around.
Focus
Data governance, data chaos
Location
Toronto, Ontario, Canada
// about

I'm a data practitioner who works at the intersection of governance, engineering, and organizational reality. The lab is where I do the work before I talk about it.

Most governance advice skips the hardest part: figuring out why the last framework failed. The answer is almost never "we lacked a policy." It's that the policy didn't connect to how decisions actually get made — who touches the data, who gets blamed when it's wrong, and who has the standing to change it.

Every experiment in this lab starts with a clear-eyed read of the real problem, commits to a principle that addresses it, and takes deliberate actions that reinforce each other. If an experiment can't be explained in those three terms, it isn't ready.

// contact

Start a conversation.

Questions, collaborations, or just curious about something you saw in the lab — reach out.

// Message received.

Thanks for reaching out. I'll be in touch shortly.

"The best governance frameworks come from people who've felt the friction."

hello@datagovernancelab.com
Toronto, Ontario, Canada
Open to conversations with practitioners, researchers, and organizations working through real governance challenges.