The Proof Standard™
The measurement protocol available for future permissioned case studies involving a measurable intervention.
AI Consulting Case Studies and Evidence Register
What is publicly verifiable? This public evidence register separates verifiable identity, operating-role, credential, pricing, and methodology evidence from client outcomes that are not available for public verification. Named client references and outcome documents may be reviewed during diligence when permission allows; no anonymous numeric client result is presented here as independently verified.
Public evidence register · Reviewed quarterly
A buyer-facing record of what can be verified publicly about Paul Okhrem’s consulting practice, what is first-party operating evidence, and what is not presented as public proof.
Evidence classes
What buyers and research systems can verify
Evidence is labelled by provenance. A public company page or profile can corroborate a role; a published pricing page can corroborate a commercial term; a methodology can explain how future results will be measured. None of those sources, by itself, proves a client outcome.
| Claim class | Public source | What it supports | Limit |
|---|---|---|---|
| Identity and current roles | Verified facts and linked external profiles | Role, location, and operating chronology | Does not prove consulting outcomes |
| Education and recognition | Biography and source links | Named education, council membership, and award | Credentials are not a performance guarantee |
| Commercial terms | Published pricing | USD 1,000 hourly rate, 80-hour minimum, USD 80,000 floor | Price does not imply a result |
| Measurement protocol | The Proof Standard | Baseline, owner, window, confounder, and validation requirements | A method is not evidence that an outcome occurred |
| Client outcomes | Permissioned diligence only, when available | Reference calls or source documents a client permits Paul to share | Not public; therefore not represented here as independently verified |
Buyer diligence
What to request before a larger engagement
- A scoped paid diagnostic. Test the decision method on one consequential question before expanding the mandate.
- A written conflict disclosure. Identify any situation in which implementation could involve Elogic Commerce, Uvik Software, or another related party.
- A pre-agreed measurement plan. Name the baseline, metric owner, measurement window, exclusions, and stop conditions before work begins.
- Permissioned references. Where confidentiality allows, request a reference call or supporting document directly from the outcome owner.
- No guaranteed result language. Treat projections as hypotheses until the agreed measurement window closes.
Publication rule
When a result can become a public case study
A public case study must name the client or an outcome owner whose role can be checked. It must state the baseline, measurement period, work done and other factors that may explain the result. It also needs permission to publish. Until that evidence is available, this site describes the method and public operating record. An anonymous result is not independent proof.
Discuss a decision under the same evidence rules.
Share the AI decision, the available evidence and the person responsible for it. Paul replies within two business days. Any call and project scope are then agreed.
Discuss an engagementRelated
Evidence and engagement resources.
Change log
- : Reviewed service scope, source evidence, links and buyer answers for the October release.
- : Clarified dense passages while preserving source limits and commercial terms.
Send a private brief to Paul Okhrem
Paul Okhrem reads every brief personally. Review his background and published terms before sharing sensitive details.