
BOOST
Business data analyst
Analyze workload and performance data to build reports and dashboards that inform leadership decisions. Recommend actions to Government leaders based on data analysis and brief them on the reasoning.

Patronus AI, Inc.
·yesterdayPatronus AI, Inc.
·yesterdayLocation
remote, United States
Salary
$302k/yr
Commitment
Part Time
Level
Middle (2-4 years)
Patronus AI is a frontier lab developing simulation research and infrastructure to accelerate progress toward human-aligned AGI. We are on a mission to simulate all of the world’s intelligence.
We are the team behind some of the earliest and most influential research in AI evaluation like FinanceBench, Lynx, SimpleSafetyTests, CopyrightCatcher, Humanity’s Last Exam, and more. We are formerly AI researchers and engineers from companies like Meta AI, Amazon AGI, and Google. Our customers include foundation model labs and Fortune 500 enterprises like Adobe. We are backed by top-tier investors like Lightspeed Venture Partners, Notable Capital, Stanford University, Noam Brown, Gokul Rajaram, and more.
You'll help us create RL tasks related to financial crime enforcement that test and train AI models.
Design realistic tasks that take an experienced engineer about a working day.
Build the sandboxed setup each task needs (transaction monitoring rules, sanctions and watchlist screening, KYC checks, case management, alerting), your own solution and checks a tester could verify without opinion.
Review tasks other experts build.
3+ years writing production code for AML, fraud, sanctions or KYC systems at a bank, fintech or vendor, managerial experience is a bonus.
Still in the role, or left recently.
Required: Previous AI data labeling, data annotation, or RL task creation experience.
Nothing copied from published sources, and no code, configs or data from a current or former employer or client. Every system and scenario is fictional.
AI tools are allowed as an aid but you are ultimately responsible and evaluated for your own work.
Patronus AI is an equal opportunity employer. We celebrate diversity in our workplace, and all qualified applicants will receive consideration for employment without regard to age, ancestry, color, family or medical care leave, gender identity or expression, genetic information, marital status, medical condition, national origin, physical or mental disability, political affiliation, protected veteran status, race, religion, sex (including pregnancy), sexual orientation, or other legally protected characteristics.
By clicking ‘Apply’, you agree to both of the following:

BOOST
Analyze workload and performance data to build reports and dashboards that inform leadership decisions. Recommend actions to Government leaders based on data analysis and brief them on the reasoning.

Xperteez Technology
Evaluate and compare original images with AI-generated versions to identify differences and visual artifacts. Document findings with clear written feedback and follow evaluation guidelines to ensure accuracy and thoroughness.

Turing
Evaluate and compare responses generated by multiple chatbots for accuracy, relevance, and completeness. Identify errors and quality issues, providing structured feedback and final assessments.

Patronus AI, Inc.
Create reinforcement learning tasks for video streaming to test and train AI models. Design realistic tasks that can be completed in a working day, and build the necessary sandboxed setups including transcoding pipelines and adaptive bitrate playback.

Patronus AI, Inc.
Design and create reinforcement learning tasks for dating apps, ensuring they are realistic and can be completed in a working day. Build the necessary sandboxed setups for tasks, including matching, recommendations, and APIs.

Patronus AI, Inc.
Create reinforcement learning tasks related to banking to test and train AI models. Design realistic tasks that can be completed in a working day, build the necessary sandboxed setups, and review tasks created by other experts.

RYZ Labs
Review and validate data matches, correcting or marking discrepancies. Label reference sets consistently and grade sampled match results against these sets. Flag image issues and submit re-imaging lists, while reporting recurring errors as patterns.