Senior Technical Product Manager, Fraud Solutions

About the role

Fingerprint's device intelligence helps businesses recognize returning devices and detect signals of potentially suspicious activity. We’re now building customer-facing solutions that turn that intelligence into something customers can put to work right away: finding abuse, understanding the evidence, taking the right action, and measuring whether it helped.

We're looking for a Senior Technical Product Manager to own our Fraud Agents solution suite. Your job is to shorten the time it takes users to leverage Fingerprint’s data in solving real-world fraud challenges. Leveraging our device intelligence primitives, sense-making, fraud scoring, rules engines, and third-party solutions, you'll shape purpose-built applications and specialist AI agents that help customers prevent problems such as trial abuse, multi-accounting, and promo or bonus abuse. Trial Abuse is the starting point. You'll help determine which problems we should tackle next and what a credible solution requires.

This is a senior individual-contributor role. You'll independently drive product direction and execution within this area, working with Engineering, Design, Data Science, other product teams, and our go-to-market partners.

What you'll own

Define solutions around customer problems

• Lead discovery with customers, prospects, fraud analysts, security teams, and developers. Understand how abuse happens, what it costs, how customers respond today, and where their existing tools fall short.
• Choose focused use cases and customer segments where Fingerprint can deliver a repeatable advantage. Turn those choices into clear product requirements, business cases, and a prioritized roadmap.
• Use our own Trial Abuse deployment and external design partners to separate reusable capabilities from customer-specific workarounds.
• Own the lifecycle from concept and prototype through alpha, beta, launch, and ongoing improvement. Establish evidence-based criteria for expanding, narrowing, or stopping an investment.

Build solutions customers can trust

• Define how an agent uses device identity, fraud signals, scores, historical activity, and other information to identify suspected abuse and recommend a response.
• Design workflows that let customers understand the evidence, review related activity, set exceptions, and correct mistakes.
• Leveraging our general purpose ML-based suspect and fraud scoring, specify how the solution can learn from confirmed abuse, legitimate-user labels, overrides, and customer policies. Keep tenant-specific knowledge separate from any shared network intelligence.
• Define autonomy levels, approval requirements, permissions, audit trails, simulation, versioning, and rollback. Decide where an agent should investigate or propose changes and where deterministic systems should execute approved policies.
• Work with Engineering and Data Science on evaluation, monitoring, and failure handling. Address hallucinated evidence, prompt injection, data poisoning, unauthorized actions, and privacy risks before they become customer problems.

Make the solution easy to adopt and operate

• Build guided onboarding, instrumentation checks, useful defaults, shadow-mode evaluation, and clear paths to enforcement. Reduce the custom engineering and ongoing tuning customers need to get value.
• Translate customer workflows into technical requirements for AI Agents, MCP Server, APIs, event schemas, account-to-device linking, historical queries, lists, integrations, and the Rules Engine.
• Identify gaps in the underlying platform and negotiate scope and sequencing with the teams that own it. Make dependencies and tradeoffs explicit.
• Partner with Design to prototype and test the operational experience, including investigation views, policy controls, explanations, and outcome reporting.
• Evaluate build, buy, and partner options when third-party enrichment or integrations would make the solution more effective.

Prove customer and business value

• Define and track success measures that balance abuse reduction with legitimate-user conversion. Account for false positives, missed abuse, analyst workload, time to value, and the cost of operating the agent.
• Establish credible ways to estimate resources or revenue protected, without treating every block as a successful outcome.
• Partner with Product Marketing, Sales, Solutions Engineering, and Customer Success on positioning, launch readiness, customer education, and adoption.
• Contribute customer evidence and recommendations for packaging and pricing. Test willingness to pay for a packaged outcome rather than assuming demand for another feature bundle.
• Write clear product requirements and decision documents. Represent the product in customer conversations, communicate risks early, and mentor less-experienced product managers.

What you bring

• A track record of independently owning a technically complex B2B product through discovery, delivery, launch, and measurable customer adoption. You can explain both the product decisions you made and the results they produced.
• A strong grasp of fraud, security, or abuse prevention. You understand attack patterns such as multi-accounting, trial and promo abuse, account takeover, credential stuffing, and malicious automation, along with the operational tradeoffs in detecting and stopping them.
• Deep understanding of how AI agents are changing fraud prevention on both sides: helping defenders investigate and respond, while making attacks cheaper, faster, and more adaptive. You can distinguish useful capabilities from unsupported claims.
• Practical fluency in agent systems, including tool use, retrieval, durable context, evaluation, human approval, and bounded autonomy. You can discuss where probabilistic reasoning belongs and where it should not control enforcement.
• Enough technical depth to work closely with engineers on APIs, SDKs, event-driven systems, identity resolution, data quality, real-time versus asynchronous processing, latency, reliability, and cost. You don't need to train models yourself, but you need to ask good questions about their behavior and limitations.
• Strong customer research and analytical skills. You can turn incomplete evidence into a testable product hypothesis, then use qualitative feedback and product data to decide what to do next.
• Sound judgment about customer trust, privacy, and false positives. You understand why shared devices, weak labels, or uncertain links should not automatically lead to a block.
• Clear writing and the ability to build alignment across teams without relying on reporting authority. You document decisions, close loops, share credit, and handle disagreement constructively.

Experience that would be especially useful

• Shipping an AI agent or AI-assisted workflow into production, with evaluation and operational safeguards.
• Building fraud decisioning, device intelligence, bot management, account protection, or trust-and-safety products.
• Turning APIs, signals, or developer infrastructure into self-service applications with repeatable onboarding.
• Working directly with fraud operations teams and with SaaS, AI, or usage-based businesses protecting trials, credits, or other entitlements.
• Designing feedback loops, investigation workflows, risk policies, or third-party data integrations.

What success looks like

Customers can connect the solution, understand its recommendations, and safely put it to work without building their own fraud operations tooling. The first Trial Abuse solution demonstrates repeatable value across customers, with measurable abuse reduction and acceptable impact on legitimate users.

We have evidence that agent-assisted investigation and policy management save customers time or catch abuse they were missing. The roadmap reflects what we've learned, and the next specialist solution reuses a proven foundation rather than adding another disconnected product.