Posted Oct 2
AI Product Manager Jobs
AI product managers and product leads decide what gets built on top of models - defining AI-native features, setting the roadmap for AI products, and working closely with research and engineering teams to turn model capability into something people can actually use.
AI Product Manager roles are rare: 37 live right now. Get the new ones every Monday.
Get new AI Product Manager jobs in your inbox
Join 100+ AI professionals · Weekly, free, unsubscribe anytime
Latest AI Product Manager roles
Posted Oct 2
Posted Oct 1
Posted Oct 1
$191K – $315K
Posted Oct 1
UPSUPS SUPPLY CHAIN SOLUTIONS, United States+3 more
Posted Sep 22
Hewlett Packard EnterpriseRoseville, California+3 more
Posted Sep 17
WorkdayUnited States+2 more
Posted Sep 17
U.S. BankMinneapolis, MN+1 more
Posted Sep 11
U.S. BankChicago, IL+1 more
Posted Sep 11
Posted Sep 10
AdobeSeattle, United States of America+3 more
Posted Sep 6
$230K – $280K
Posted Sep 3
Procter & GambleCINCINNATI GENERAL
Posted Sep 1
Posted Aug 19
Posted Aug 12
Posted Aug 12
Posted Aug 9
Posted Aug 9
Posted Aug 9
$205,600 - $257,000
Posted Aug 4
$205,600 - $257,000
Posted Aug 4
$154,084.36 - $263,131.83
Posted Jul 23
Posted Jun 30
$177,000 - $220,000
Posted Jun 25
$100,000 - $200,000
Posted Jun 19
Posted Jun 17
$180K – $225K
Posted Jun 12
Posted May 21
Posted Apr 13
$200,000 - $250,000
Posted Mar 3
Posted Jan 13
Posted Dec 12
Posted Nov 20
Posted Oct 12
$205,000 - $257,000
Posted Sep 23
$171,200 - $214,000
Posted Sep 18
When the eval set is the spec
AI product management breaks the central artifact of the discipline: you cannot write a deterministic specification for a probabilistic system. There is no acceptance criterion of the form 'given this input, produce exactly this output' when the same input yields different outputs. In practice the evaluation set becomes the spec, a curated collection of representative cases with graded expectations, and the product manager usually owns its contents, because deciding what good output looks like is a product judgment before it is an engineering one.
Inference cost lands in cost of goods sold, which makes unit economics a product problem rather than an infrastructure footnote. Every feature decision, from how much context to send, to which model tier to route to, to whether an agent may take another step, carries a direct marginal cost per use, and a generous free tier can lose money on each active user. AI product managers are expected to reason about cost per request alongside adoption and retention in a way that most software product managers never had to.
Two flavors hire separately and do not interchange cleanly. Platform and API product management serves developers: model and endpoint capabilities, rate limits, pricing and packaging, documentation, and developer experience. Consumer and application product management serves end users: interaction design for uncertain output, disclosure and trust, error recovery, and habit formation. The strengths that make someone excellent in one are not the ones that get them hired into the other.
Explore related searches
- Browse the AI Engineer hub for the full specialization.
- Prefer remote? See remote AI jobs.
- AI Governance jobs
Frequently asked questions
- How many AI Product Manager positions are listed at the moment?
- We are tracking 37 live AI Product Manager roles across the AI companies we monitor, updated hourly. Each listing links straight to the employer's own application page.
- How do you write a product spec for a non-deterministic feature?
- By specifying acceptance in terms of evaluation rather than exact output. Product teams define representative inputs, grading rubrics for what counts as a good response, thresholds required to ship, and explicit behavior for failure and uncertainty. The evaluation set becomes the living specification, and product managers typically own its contents even when engineers own the harness.
- Why is inference cost a product management concern?
- Because it is a marginal cost incurred every time the feature is used, so it sits in cost of goods sold rather than in fixed infrastructure spend. Context length, model tier, retry policy, and agent step limits are all product decisions with direct cost consequences, and pricing or packaging that ignores them can lose money as usage grows.
- What is the difference between platform and application AI product roles?
- Platform and API roles serve developers building on models, covering capability exposure, rate limits, pricing and packaging, documentation, and developer experience. Application roles serve end users, covering interface design for uncertain output, disclosure and trust, correction and recovery flows, and adoption. Employers hire for them separately and rarely treat the experience as interchangeable.