Anthropic’s Account Abuse team seeks a Staff+ Software Engineer to build and operate machine learning systems that detect and mitigate account abuse and fraud at scale. The role involves full‑stack ML engineering, from feature platform development to model training, evaluation, and deployment, leveraging Claude and other tools to accelerate the model lifecycle. Candidates will collaborate closely with data scientists, policy, and product teams to ensure robust, precise, and safe solutions.
Staff+ Software Engineer, Account Abuse (Machine Learning) na Anthropic
Híbrido - San Francisco, CA
Ver mais vagas na AnthropicSalary
USD 320,000 - 485,000
Requirements
Skills
- Proficiency in Python and SQL
- Experience training machine learning models and deploying them to production
- Experience building data pipelines with a batch processing engine (e.g., Spark, Beam) and a workflow scheduler (e.g., Airflow)
- Working understanding of point‑in‑time correctness and training / serving skew, and how to prevent both
- Strong communication skills and ability to explain technical tradeoffs to non‑technical stakeholders
- Experience building or operating a feature platform such as Chronon, Feast, or Tecton
- Experience with stream processing engines such as Flink, Beam / Dataflow, or Kafka Streams
- Experience training ML models in a production setting with demanding serving requirements, such as fraud, risk, or ranking
- Experience with tree‑based models on tabular data
- Experience building unsupervised, clustering‑based or graph‑based detection systems to surface coordinated account abuse
- Experience in integrity, spam, fraud, or abuse detection
- Experience working with scarce, delayed, or noisy labels
- Experience with AutoML or other approaches to automating the ML workflow
- Care about the societal impacts of AI and want your work to make powerful systems safer
- Bachelor’s degree or equivalent combination of education, training, and/or experience
Responsibilities
- Build and operate a feature computation platform that serves both model training and real‑time scoring, with point‑in‑time correct training data and low‑latency online retrieval
- Train, evaluate, and deploy models that detect account‑level abuse and fraud, running them both offline and online
- Build tooling that automates more of the model development lifecycle, including using Claude to speed up feature development, training, and evaluation
- Make backtesting, shadow deployment, and staged rollout the default path to production, with monitoring for training / serving skew, drift, and adversarial adaptation
- Work with our data scientists and our Policy & Enforcement team to improve label coverage and quality
- Partner with product and platform teams to gather signals and integrate model decisions with minimal impact on their systems’ latency, stability, or overall architecture
Technologies
PythonSQLSparkBeamAirflowChrononFeastTectonFlinkKafka StreamsAutoMLClaude
Descubra se seu currículo está pronto para esta vaga
Veja como nossa IA pode otimizar seu currículo e aumentar suas chances de conseguir esta posição.