AINPM.in / Skills
Proof of skill · Updated 5 Nov 2026

What I can do, and the evidence you can check

Eight AI-native product management skills, written the way employers describe them, each linked to public evidence: demos, code, evaluation results, publications and credentials. A skill claim is only as strong as the proof behind it.

44%of learners want a clearer description of the skills they gainedAccredible, 2026
22→44%of issuers now rank proving program value as a top priorityAccredible, 2026
75%of HR leaders say digital credentials make skills easier to verifyAccredible, 2025
49evidence links on this page, each one click awayAINPM.in

Accredible’s 2026 figures come from spring 2026 surveys of 1,000 learners and 194 credential issuers. Accredible sells credentialing software, so treat its numbers as directional. Learn more in my analysis of the research

At a glance

Where is the proof strongest?

The heatmap counts the public evidence behind each skill by type. The map shows which products and writing provide it. Both are generated from the matrix below, so they always match.

Evidence by skill and type49 public evidence links · darker = more evidence · select a cell to filter the matrix
Number of public evidence links for each skill, by evidence type
SkillLive demoOpen codePackageContributionEval resultsPublicationCredentialCase studyTotal
AI evaluation design···122128
Retrieval (RAG) product design23····117
Trust and compliance by design (India)1···13·16
Privacy and security architecture·12··1·15
On-device and local AI13···1·27
AI product strategy and MVP scoping·····3126
Technical program delivery··11··215
Communicating AI quality honestly···112·15
Total473341251149
Which projects prove which skills?29 connections · hover or focus a skill or project to trace it
Map of eight skills on the left connected to the projects, writing and credentials that provide evidence for them, on the right. The same information is listed below the chart.AI evaluation designRetrieval (RAG) product designTrust and compliance by design (India)Privacy and security architectureOn-device and local AIAI product strategy and MVP scopingTechnical program deliveryCommunicating AI quality honestlySEBI Circular RAG5 skillsTrustVault2 skillswebauthn-prf-zktv2 skillsLecture Lens3 skillsCodeSearchNet4 skillsOpen-source contributions4 skillsArticles & guides5 skillsCredentials4 skills
Show as a list
  • SEBI Circular RAG: AI evaluation design, Retrieval (RAG) product design, Trust and compliance by design (India), AI product strategy and MVP scoping, Communicating AI quality honestly
  • TrustVault: Privacy and security architecture, AI product strategy and MVP scoping
  • webauthn-prf-zktv: Privacy and security architecture, AI product strategy and MVP scoping
  • Lecture Lens: AI evaluation design, On-device and local AI, AI product strategy and MVP scoping
  • CodeSearchNet: AI evaluation design, Retrieval (RAG) product design, AI product strategy and MVP scoping, Communicating AI quality honestly
  • Open-source contributions: AI evaluation design, Privacy and security architecture, Technical program delivery, Communicating AI quality honestly
  • Articles & guides: AI evaluation design, Trust and compliance by design (India), On-device and local AI, AI product strategy and MVP scoping, Communicating AI quality honestly
  • Credentials: AI evaluation design, Retrieval (RAG) product design, AI product strategy and MVP scoping, Technical program delivery
01 · Skills matrix

Which skills, and where is the proof?

Each skill is written as one sentence an employer would recognise, aligned to an O*NET work activity and a NACE career readiness competency, then backed by evidence anyone can open.

  1. AI evaluation design

    Turn a product requirement into a graded test set with metric floors that block a release.

    Proof260-query adjudicated golden set; abstention accuracy 0.981 and recall@10 0.943 on SEBI Circular RAG.

    Framework mapping · NACE: Critical Thinking
    O*NET work activity
    Evaluating Information to Determine Compliance with Standards · Analyzing Data or Information
    NACE competency
    Critical Thinking
  2. Retrieval (RAG) product design

    Design retrieval products that cite sources, track which documents are still in force and refuse when evidence is missing.

    Proof1,490 SEBI circulars indexed with supersession-aware citations and as-of-date answers.

    Framework mapping · NACE: Technology
    O*NET work activity
    Getting Information · Processing Information
    NACE competency
    Technology
  3. Trust and compliance by design (India)

    Translate DPDP, RBI FREE-AI, SEBI and MeitY obligations into product controls in the first sprint.

    ProofRegulation radar with dated obligations; abstention and citation controls shipped in a regulated-domain product.

    Framework mapping · NACE: Professionalism
    O*NET work activity
    Evaluating Information to Determine Compliance with Standards · Monitoring Processes, Materials, or Surroundings
    NACE competency
    Professionalism
  4. Privacy and security architecture

    Specify zero-knowledge storage and passkey-based key handling, with a written threat model.

    ProofPublished npm library, a DOI-registered preprint and a PWA with 4,139 passing tests (6 Jul 2026).

    Framework mapping · NACE: Technology
    O*NET work activity
    Working with Computers · Thinking Creatively
    NACE competency
    Technology
  5. On-device and local AI

    Ship AI features that run on the user’s device so personal data never leaves it.

    ProofLecture Lens runs transcription, notes and flashcards entirely on a Mac; TrustVault explains password risk with on-device models.

    Framework mapping · NACE: Technology
    O*NET work activity
    Working with Computers · Updating and Using Relevant Knowledge
    NACE competency
    Technology
  6. AI product strategy and MVP scoping

    Frame the user job, pick the smallest product that proves value and decide what not to build.

    ProofFour open-source products taken from idea to a public demo, package or staging app.

    Framework mapping · NACE: Leadership
    O*NET work activity
    Developing Objectives and Strategies · Making Decisions and Solving Problems
    NACE competency
    Leadership
  7. Technical program delivery

    Run cross-team delivery of cloud-native platforms with clear scope, risk and release control.

    ProofDay-to-day work on mutual fund AMC products and platforms, backed by PMP (verifiable on Credly) and CSM (certificate on request).

    Framework mapping · NACE: Teamwork
    O*NET work activity
    Coordinating the Work and Activities of Others · Scheduling Work and Activities
    NACE competency
    Teamwork
  8. Communicating AI quality honestly

    Explain what an AI product does well and badly, in numbers, to non-specialists.

    ProofCase studies publish weak metrics too, such as citation precision 0.194, alongside rejected experiments.

    Framework mapping · NACE: Communication
    O*NET work activity
    Communicating with People Outside the Organization · Interpreting the Meaning of Information for Others
    NACE competency
    Communication

See the skills passport: ESCO and SFIA 9 mapping with levels

Framework labels are my own alignment for orientation. They are not an endorsement or certification by O*NET, NACE or any framework owner.

02 · Credentials

Which credentials back these skills?

Credentials are one signal, not the whole case. Accredible's 2025 research found 84% of HR leaders have suspected or found exaggerated qualifications, and 53% verify claims frequently. So each credential below links to the issuer's own verification page where one exists, shows its dates, and names the skills it supports. Records I uploaded myself are labelled as such.

7 of 12 credentials are verifiable on the issuer's platform · checked 5 Nov 2026

certification

Certified ScrumMaster (CSM)

Scrum Alliance

Certificate available on request

Supports: Technical program delivery

No public verification link
03 · Method

How can you build your own proof-of-skill matrix?

The same five steps work for any AI product manager in India, whether you are moving into AI from a traditional PM role or applying to a capability centre.

Name five to eight skills for the role you want

Start from real job descriptions, not a course syllabus.

Write each as one employer-readable sentence

“Turn a requirement into a graded test set” beats “strong AI skills”.

Map it to a recognised framework

O*NET, ESCO or NACE help recruiters and AI search tools match your skills to roles.

Link at least one public artefact per skill

A demo, repository, eval log, write-up or verifiable credential. Publish weak results too.

Review it every quarter

Add new evidence, retire stale claims, and date the page so readers know it is current.

04 · FAQ

Questions about skills and evidence

What is a proof-of-skill matrix?

A proof-of-skill matrix lists each skill in language employers already use, then links to public evidence for it, such as a live demo, source code, evaluation results or a publication. It lets a hiring manager or client check a claim in minutes instead of trusting a list of certificates.

Why map skills to O*NET and NACE?

Accredible’s 2026 State of Credentialing research found that 44% of learners want a clearer description of the skills they gained, and recommends describing skills in frameworks employers already use. O*NET work activities and NACE career readiness competencies are widely used by employers and are two of the frameworks Accredible added in August 2026.

Are certificates still useful for AI product managers?

Yes, as one signal among several. Accredible’s 2025 research found 75% of HR and talent leaders say a digital credential makes skills easier to verify, but its 2026 report concludes that credentials alone do not guarantee career outcomes. Pair each credential with shipped work and measured results.

How can I build my own proof-of-skill matrix?

List five to eight skills for the role you want, write each as one sentence an employer would recognise, map it to a framework such as O*NET or NACE, and link at least one public artefact per skill: a demo, repository, evaluation log, write-up or verifiable credential. Publish weak results too.