/roles — ROLE_4
Machine Learning Engineer
They are the leading agentic risk platform to fight financial crime
The role
- COMP
- $175K - $220K
- EQUITY
- Competitive equity
- LOCATION
- New York · San Francisco · South Bay Area · Los Angeles · Boston · Seattle · Texas · Chicago · Washington DC · Denver · Florida · Minnesota · Sacramento
- WORKPLACE
- Remote
- EXPERIENCE
- 5 - 8 years
- VISA
- None
- STACK
- Go, Python, PyTorch, Scikit-learn, Docker, Kubernetes, CI/CD
- INDUSTRY
- Financial Services, Fintech, Cybersecurity, Security
The company
Agentic financial-crime platform used by leading banks and merchants worldwide to stop fraud in real time and automate fraud and AML operations.
- STAGE
- scale-up
- FUNDING
- $170M+ raised
- TEAM
- 200+ people
- FOUNDED
- 2019
- BACKING
- backed by a16z and major financial-industry strategics
JD — the work
About the role
More than a modeling job: you'd own the full path from raw device and behavioral signals to live fraud decisions — the models, the data pipelines, and the Go backend that keeps everything fast and reliable at production scale, in a high-stakes and constantly shifting adversarial domain.
What you'll do
- Build and tune real-time pipelines and backend services for device and behavioral data
- Develop, deploy, and maintain fraud-detection models that hold up in production
- Turn raw signals into production-ready features
- Integrate tightly with platform and backend engineering
- Keep security, privacy, and compliance standards high
- Champion testing, documentation, and observability
What they're looking for
- 5+ years of software engineering with strong backend depth (Go or Python)
- Hands-on applied ML on large datasets (PyTorch, scikit-learn)
- Strong SQL across relational and non-relational stores
- End-to-end ML system experience: feature pipelines, deployment, monitoring
- BS/MS in CS, engineering, or related field
Nice to have
- Fraud, risk, or cybersecurity domain knowledge
- CI/CD, Docker, Kubernetes fluency
- Modern browser APIs and high-entropy data collection
- Using frontier LLMs for automation