SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Signifyd is hiring a Machine Learning Engineering Manager to lead one of the teams within Signifyd AI Lab (SAIL), the department responsible for building ML products that power fraud and risk decisions across e-commerce transactions at scale. The role is a hands-on player-coach position in a department that balances near-term continuous model improvements with longer-horizon innovation bets.
As manager, you will lead and grow a geographically distributed team, guiding career development, mentoring engineers, managing conflicts, and fostering a collaborative environment. You'll engage in regular 1:1s, provide constant feedback, and create psychological safety—especially around experiments that don't succeed. You'll identify capability gaps and enhance team efficiency through technical guidance and support in both hard and soft skills.
You will run a portfolio of experiments rather than a delivery queue. You'll partner with tech leads who own the technical roadmap, pressure-testing ideas, sharpening them, and ensuring the strongest hypotheses get resourced. You'll make critical trade-off decisions: which hypotheses receive compute and headcount, which get another iteration, and which receive a clear documented "no." You'll manage the balance between committed improvement targets and research bets with longer payoff horizons, re-cutting budgets as evidence arrives and explaining reasoning to both team and stakeholders.
Rigor is essential: you'll ensure offline results predict online behavior, treat strong point estimates as starting points rather than conclusions, and own delivery on a cadence where independent workstreams converge into release candidates, get evaluated end-to-end, and ship—including the difficult decision to exclude workstreams that aren't delivering value.
You'll set direction in partnership with Risk stakeholders, your primary customers. Commitments are explicit, measured, and documented; you deliver model performance improvements while Risk owns thresholds, rules, and application of decisions.
The ideal candidate brings technical credibility—staying close enough to code and evaluation pipelines to have grounded opinions and distinguish statistical results that hold in production from those that merely look good in test windows. You balance a high bar for evidence without becoming a bottleneck to experimentation. You possess executive judgment: balancing research bets against quarterly delivery, disagreeing and committing when decisions are made, and building an environment where well-documented negative results are celebrated as real progress.
