Sr. Staff AI/ML Engineer: Kraken

This is not a "Data Scientist" role where you build models in a notebook. This is a Platform Architecture role. You are building the "Machine Learning Factory" that the rest of the company will use.

As a Senior Staff Engineer, you are expected to operate "above the code." While you will still be hands-on, your primary output is Technical Strategy and Standardization. You are solving the "N+1" problem: How do we make it easy for the next 50 engineers to deploy models without breaking production?

Key Responsibilities
The Architect: You own the design of the training, serving, and monitoring infrastructure. You decide: "Do we use Ray or Spark? feature store or ad-hoc pipelines?"

The Standardizer: Currently, different teams (Risk, Marketing, Product) might be doing ML differently. Your job is to unify them under one reliable framework.

The High-Stakes Engineer: You are building systems for Fraud and Risk. In crypto, if a recommendation engine fails, users get annoyed. If a fraud model fails, Kraken loses millions of dollars in seconds. Reliability is paramount.

GenAI Adoption: You are the gatekeeper for new tech. You determine if a GenAI tool provides "durable value" or if it's just hype.

Strategic Analysis: The "Crypto-Native" ML Stack
High Performance Requirement: The JD mentions Rust and Go. This is rare for typical ML roles (usually just Python) but common in Crypto. They likely need low-latency inference for high-frequency trading or real-time fraud detection.

The "Sr. Staff" Expectation: At this level (L6/L7 equivalent), you aren't just given tasks. You are expected to find the problems. You must partner with "Product and Data Leaders" to figure out where ML can actually make money.

The "Merit" Culture: Kraken is famous for its explicit "Libertarian/Cypherpunk" culture code (which they link to in the JD). They value "High Agency" and thick skin. If you are sensitive to direct feedback or rigorous debate, this culture might be a shock.

Candidate Profile
The "Plumber": You have 10+ years of experience, but largely in Infrastructure. You know Kubernetes, Kubeflow, MLflow, and Airflow better than you know the math behind a Transformer model.

The Polyglot: You are an expert in Python, but you are not afraid to write high-performance services in Rust or Go.

The Scaler: You have worked at a company where the data volume is massive (Petabytes). Experience with Spark and Distributed Systems is a hard requirement.

Critical "Knockout" Criteria
10+ Years Experience: They are strict on seniority.

Platform vs. Product: If your resume only shows "Improved accuracy of model X by 2%," you will be rejected. They want to see "Built the platform that allowed 50 models to be deployed."

System Design: Can you design a system that handles 100,000 transactions per second with <50ms latency?