Senior Product Manager, RLHF

Insightful shows companies how work actually gets done. Our Work Intelligence Platform turns millions of work data events into Workforce Analytics, People Intelligence, and Process Intelligence.

The human data companies that train frontier AI models run some of the largest expert networks in the world, and they pay those experts for every hour. You’ll own the next version of our product for this market, and the job has two parts. First, adapt our Workforce Analytics product for RLHF: a targeted, API-first version that customers plug into their own platforms. Second, build fraud intelligence on top of it: what it detects, what a trust score means and what we are willing to vouch for, and how the product moves from tracking hours to tracking the work itself. You need to be equally strong in product strategy, APIs, and AI.

This is one of our fastest-growing segments. We have an established footprint, customers already building on our data, and plenty of room to grow the product offering. The role is a senior individual contributor seat today, with a direct path to running the business unit.

This is a hybrid role based in New York City. You’ll be in the office at least two days a week (Tuesdays and Thursdays).

• →Know the industry inside and out. How contracts are won and lost, how experts are sourced and vetted, how work gets scoped, reviewed and paid for, and where they lose money. Customers should ask you what the rest of the market is doing.
• →Adapt the product for RLHF. Turn Workforce Analytics into what a human data company needs and make it API-first: what stays, what goes, and the endpoints, schemas and docs their engineers build on.
• →Build fraud intelligence. Define what counts as fraud from the customer’s side (AI-assisted work, padded time, shared accounts, scripted activity), what an expert’s trust score means, and what we stand behind.
• →Build the product forward. Own the roadmap past the first version. Decide what customers need next and push the industry forward.

• Five or more years as a PM on technical products, including an API or platform that customers built on
• Hands-on understanding of how RLHF works: how experts produce annotation and preference data, how it gets reviewed and paid for, and where the incentive to cheat sits
• Shipped something with a model in the loop and owned the eval that said whether it worked
• Product sense in ambiguity: you know what the customer needs before they say it, and you turn complex ideas into simple solutions
• The ambition to run a business, not a backlog: you want ownership of a product area and eventually a business unit
  

• Built or run product for an expert network, contractor or gig workforce at a scale where fraud was a line item: onboarding, quality control, payment, offboarding
• Worked at a human data company or expert network
• Fraud, risk, or trust and safety background in any domain