Staff Product Manager, Risk Modeling

Plaid•San Francisco, CA

About The Position

We protect the integrity of Plaid's network and authentication surfaces. We keep bad actors, synthetic and stolen identities, and low-quality data-partner traffic off the platform so that the millions of connections Plaid powers stay trustworthy. We build Plaid's internal risk platform: detection, scoring, and real-time decisioning and we're the foundation every other Plaid product depends on. You own Plaid's internal network-health and risk platform — how we detect, score, and act on risky traffic across our 10,000+ data-partner integrations. You set the roadmap, define the block/step-up/allow decisioning framework, and partner directly with the Data Science and MLE and operations team to sharpen detection precision. This is a platform PM role: your customer is the network itself, and your output is a measurably safer, higher-quality data supply.

Requirements

6+ years in product management, ideally on internal risk, trust & safety, fraud, or identity platforms.
Hands-on technical depth: detection signals, precision/recall tradeoffs, light data analysis/investigation.
Experience building ML based products / insights
Proven partnership with Data Science / MLE teams on risk models.
Experience with risk decisioning, KYC, synthetic/stolen-identity detection, or ATO.
Clear product requirements; strong cross-functional alignment.

Nice To Haves

Experience in a sophisticated internal risk org at a large fintech/bank/marketplace.
Former risk analyst / DS / investigator who moved into PM; familiarity with bank data / open finance.

Responsibilities

Define and prioritize the DS/MLE roadmap for network-health detection — articulate model investment needs, own precision/recall tradeoff decisions, and translate model outputs into real-time product actions.
Build and own the data-driven framework for real-time block-vs-allow decisions, balancing network quality against customer and commercial impact.
Aggregate requirements across risk engineering, fraud & abuse operations, and data partners into a single prioritized roadmap.
Define how to measure Network Health.
You define and own the team’s OKRs (risky-traffic detection/step-up; ATO reduction, Authentication correctness).
Define and recommend Plaid’s risk-versus-conversion posture, partnering with company leadership to set clear, data-backed thresholds for when to allow, step up, or block traffic.
Get hands-on with data: run light investigations yourself, pull signals, and ground product decisions in evidence — this PM doesn't wait for an analyst to tell them what's happening.
Span multiple risk problem categories simultaneously: data-partner supply risk, authentication risk/correctness, and AQO — owning one coherent roadmap across all three.