Eighteen CDO-grade diagnostics that run inside your AI assistant, plus the practitioner who built them. Know where you stand, build it right, run it in production, and prove the value.
Not sure where to begin? See where to start →
A scored, comparable baseline of your data, your AI readiness and the organisation that has to act on it.
See Data & AI Maturity Assessment →
Score the data estate and its AI readiness the same way, every time.
Decide who owns what, then check the organisation can actually execute.
Architecture, master data, definitions, vendors, deals and estimates, decided with a method you can defend.
See Data & AI Architecture →
Choose the pattern and the golden-record style before the spend is committed.
One meaning for every term that ends up on a board report.
Select vendors, diligence a deal and price the work without guessing.
Quality, lineage, migration, test data and AI behaviour: the operational layer where programmes quietly succeed or fail.
See Data Lineage & Audit Evidence →
Route every defect to an owner, and prove where every number came from.
Move data safely between systems and environments.
Plan the first hundred days of an agent, control what it remembers, and evidence where it is biased.
Put a defensible number on bad data, platform spend and AI output, from your own evidence.
See Cost of Bad Data →
What bad data and AI errors actually cost, in dollars Finance accepts.
Split shared platform costs by what each team actually uses.
Every role asks a different first question. These are the skills each one usually starts with.
Baseline the estate, deploy the org, fund the fix.
Architecture, migration, test data and platform spend.
What bad data, cloud and AI actually cost, and what they return.
Evidence that survives an examiner, and AI you can defend.
Know what you're buying, and what separation costs.
Each skill packages a consulting method into a structured conversation inside your AI assistant. It asks the questions, scores the answers the same way every time, and produces a finished output.
Start from what the board is asking: where are we, what is it costing, can we trust this number.
One sitting with the people who know the answers. No data extract, no tool install, no procurement cycle.
Take the scored output and red flags into a working session with Harveer, or run with them yourself and re-run next quarter.
The skills themselves are not downloadable from this site. They live in private repositories on GitHub, and Harveer grants access personally, usually after a working session.
A full DataPulse assessment (3–5 hours), against a typical 12-week consulting assessment.
Of structured analysis the eighteen skills stand in for.
Fixed dimensions and scoring, so the second run shows movement.
The skill does the analysis. You still own the decisions.
Consulting comparisons are practitioner estimates (team size × weeks for the same analysis bought as an engagement). Firms don't publish rate cards, so there is no citable public benchmark. They are offered to be challenged.
Fortune 500 Chief Data Officer and author of When Data Moves. Every method here comes from building and running the thing, not recommending it.
Thirty minutes. Bring the question your board is asking.
Thirty minutes with Harveer. Tell him what your board is asking, and he'll tell you which skill to run first, and whether you need anyone alongside it.