AI Delivery Professional · AIDP
Most enterprise AI never ships. AIDP certifies the people who make it deliver.
A single four-hour examination of delivery judgment, conferred by ADaSci. Earned, never attended. Valid for two years.
- $350
- per sitting
- 4 hrs
- one sitting
- ~35%
- indicative pass rate
- 13 Nov
- next sitting
The first holders
They sat it. They passed.
Founding holder
Amar Ramdhave, AIDP
AI Delivery & Quality Engineering Leader
Independent Advisor (formerly Director of Quality Engineering, Coupa Software)
Founding holder
- “AI rarely fails on the model. It fails on the decisions around it.” — Amar Ramdhave
- “Strategy makes the headline; execution makes it real. Delivery is the real differentiator.” — Aditya Vidwans
- “Most AI that never gets adopted was never framed around a real problem in the first place.” — Amar Ramdhave
- “A digital transformation programme completed on time and on budget, but not adopted by its intended users, may still be a failure.” — Aditya Vidwans
- “Lead from outcomes, not activity.” — Amar Ramdhave
- “The exam itself can be a huge learning experience.” — Aditya Vidwans
01 · The role
Engineers build it. Risk teams govern it. Someone has to deliver it.
The AI Delivery Professional is accountable for the outcome: the person the room looks at when a working model has not changed a single business number.
Engineer
Builds it
Ships the model, the pipeline and the interface. Measured on whether it works.
Risk & compliance
Governs it
Holds the line on data, policy and model risk. Measured on whether it is safe.
AI Delivery Professional
Delivers it
Accountable for the outcome: value landing before budget does, inside the workflow that already exists.
- 01Frame the decision that justifies the project.
- 02Choose what to build, buy, or ignore.
- 03Sequence delivery so value lands first.
- 04Hold the line on data, risk and reversibility.
- 05Engineer adoption inside the existing workflow.
02 · What is examined
Six domains of judgment, weighted.
The competency blueprint is authored by AIM Research and fixed by an examination board. Every item is a scenario, and every domain has a known weight.
- Opportunity Framing & Value Definition18%
- Feasibility & Solution Decisions16%
- Delivery Planning & Execution18%
- Data, Risk & Responsible Deployment18%
- Adoption, Change & Value Realization16%
- Stakeholder, Cost & Portfolio Leadership14%
03 · The exam
Built to be hard. On purpose.
If everyone passes, nothing has been certified. The cut score is fixed by a standard-setting panel, not by a curve, and we expect around one in three candidates to clear it.
- 4 hrs
- one sitting, computer-based
- Scenario
- item sets, not flashcards
- Best answer
- judgment items, four options
- In person
- proctored at the sitting city
Indicative pass rate
~35%of candidates clear the bar on first attempt
The designation self-selects for candidates whose judgment holds under load. Retakes are allowed at the next sitting.
Sample item · The real shape
A health insurer's claims-triage model is 91% accurate. Six months in, only 6% of claims are routed through it and cycle times have not moved. The recommendation lives in a separate dashboard; adjudicators are penalised for wrong approvals, not for slowness.
What is the highest-leverage first action?
Sit the full sample exam04 · Who sits it
For the practitioner who is accountable for the outcome.
Registration is open and the exam self-selects. We recommend five or more years delivering or leading enterprise AI and data initiatives.
- Product managers
- Delivery leads
- Solution architects
- Transformation leads
- Consultants
- Heads of AI & data
- Program owners
Experience · 0 yrs5+ yrs recommended
05 · Fees & what is included
One fee. One attempt. Everything you need to sit.
Examination fee
$350USD · one attempt
- Examination sitting, one attempt
- Listing on the AIDP holder registry
- Post-nominal and holder mark
- Designation conferral on passing
- Candidate Handbook and Competency Blueprint
- Two-year validity
“Most AI that never gets adopted was never framed around a real problem in the first place.”
06 · The standard
Authored, governed, and fixed.
Never guessed.
Register
Choose a sitting. Registration is open; the exam self-selects.
Prepare
Two documents to read. You do not study your way to this.
Sit
Four hours, in person, computer-based, one session.
Cut score
Fixed by the standard-setting panel, never by a curve.
Conferred
Use AIDP after your name. Renew by examination in two years.
Authored
Body of knowledge by AIM Research
Six domains, weighted, published in full in the Competency Blueprint.
Governed
An examination board owns the standard
Item writing, review and standard setting sit with the board, not the commercial team.
Set, not guessed
A defensible bar of competence
The cut score is fixed by an expert panel. By design, most do not clear it on the first attempt.
07 · FAQ
The short answers.
Everything else is in the Candidate Handbook and the Competency Blueprint, both free to download.
No. It is a single examination. There is no curriculum to complete and no attendance requirement. The Candidate Handbook and the Competency Blueprint are the only two documents you need.
Practitioners accountable for whether enterprise AI delivers an outcome: product managers, delivery leads, solution architects, transformation leads, consultants, heads of AI and data, and program owners. Five or more years of experience is recommended, not mandatory.
Deliberately hard. Four hours of scenario-based, best-answer items, sat in person. The cut score is fixed by an expert panel, and the indicative first-attempt pass rate is around 35%.
USD 350 per sitting, covering one attempt. Renewal is USD 150 every two years.
Singapore and New York on 13 November 2026, Sydney on 20 November and Austin on 11 December. In 2027, Bengaluru and Hong Kong on 26 February and Hyderabad on 5 March.
The AI Delivery Professional designation, conferred by ADaSci: the right to use AIDP after your name, a certificate, a digital badge and a public verification page for your credential number.
Next sitting: 13 November 2026.
Singapore and New York, then Sydney and Austin. $350 per sitting. Read the Candidate Handbook before you register.
Explore more about AIDP