ML Engineer

Role overview

Docker is hiring a ML Engineer as one of the founding engineers for the Intelligence team (Intelligence Org). You’ll work with the team’s first engineers and manager to decide what to build, how to build it, and how it fits into the broader Docker platform. This is a hands-on builder role with staff-level scope: you’ll shape technical direction and ship the first intelligence capabilities to customers.

Responsibilities

• Design, train, evaluate, and ship ML systems for governance and security use cases, including:
  • Prompt injection detection
  • Behavioral anomaly detection
  • Trust scoring
  • Policy recommendations
• Build and own ML infrastructure, such as:
  • Data pipelines and feature stores
  • Model serving
  • Evaluation harnesses
  • Feedback loops to accelerate iteration
• Make pragmatic build-vs-buy decisions using frontier models, off-the-shelf tooling, and managed services; build custom systems when they create durable advantage.
• Set the technical direction for the team’s ML work:
  • Own architecture
  • Define evaluation methodology
  • Manage model lifecycle and shipping bar
• Recruit and mentor as the team grows.
• Participate in a potential 24/7 on-call rotation for agentic platform services you build/operate (pager responsibility).

Requirements

• 5+ years deep applied ML/AI expertise with a track record of shipping production ML systems.
• Experience in fraud, abuse, safety, security, or trust domains (adversarial dynamics, imbalanced data, high-stakes decisions).
• 4+ years hands-on full-time software engineering experience in backend/infrastructure/platform engineering.
• Bachelor’s degree in CS/Engineering (or equivalent practical experience).
• Built and owned end-to-end ML systems around data pipelines, serving, evaluation, monitoring, shipping customer-facing products.
• Fluency with modern AI tools, including strong judgment on when to use frontier models vs traditional ML.
• Production LLM systems experience: evaluation, prompt engineering, fine-tuning, retrieval, guardrails, and agent frameworks.
• Familiarity with the agent / MCP ecosystem.
• Enjoys early-stage work where the roadmap evolves while building; makes clear decisions with incomplete information.
• Collaborative, low-ego working style.

Nice-to-haves

• Not explicitly listed, but the agent/MCP ecosystem familiarity and deep trust/security domain experience are strongly emphasized.