Staff ML Engineer: Account Abuse & Fraud Detection

Anthropic Limited; San Francisco (CA); Hybrid; USD 320,000 - 485,000; Full time

Benefits offered by this job

Equity donation matching

Generous vacation and parental leave

Flexible working hours

Office in San Francisco

Job summary

Anthropic a leading AI safety company in San Francisco, is seeking a software engineer on the Account Abuse team. You will build machine learning systems to detect and stop abuse at scale, focusing on productionizing models and low-latency real-time scoring.

The role emphasizes understanding adversarial behavior, measuring precision, and safe rollout, with collaboration across data scientists, policy, and product teams.

Qualifications

• Bachelor’s degree or equivalent in a related field
• Strong Python and SQL skills
• Proven experience training and deploying ML models to production
• Experience building data pipelines with Spark, Beam, Airflow
• Understanding of point-in-time correctness and data skew
• Excellent communication of technical tradeoffs to non-technical stakeholders

Responsibilities

• Build and operate a feature computation platform for model training and real-time scoring
• Train, evaluate, and deploy models that detect account abuse and fraud
• Develop tooling to speed up feature development, training, and evaluation
• Implement backtesting, shadow deployment, and staged rollouts with monitoring
• Collaborate with data scientists and policy teams to improve label quality
• Work with product and platform teams to minimize latency impact of model decisions

Skills

Python

SQL

ML model training

Data pipelines

Point-in-time correctness

Technical communication

Education

Bachelor’s degree in a relevant field

Tools

Spark

Beam

Airflow

Flink

Kafka Streams

Job description

Anthropic a leading AI safety company in San Francisco, is seeking a software engineer on the Account Abuse team. You will build machine learning systems to detect and stop abuse at scale, focusing on productionizing models and low-latency real-time scoring.

The role emphasizes understanding adversarial behavior, measuring precision, and safe rollout, with collaboration across data scientists, policy, and product teams.