Principal Machine Learning Engineer, Engineering Acceleration

Roblox Roblox · Consumer · San Mateo, CA · Machine Learning

Principal ML Engineer to architect the Reasoning Layer and design systems for AI agents to navigate a billion-line codebase. Focus on AI workload optimization, synthetic data generation, and agent evaluations. Requires expertise beyond basic fine-tuning, focusing on RLCF and Domain-Specific Distillation for code-specific optimization and continuous learning loops.

What you'd actually do

  1. Architect the Reasoning Layer and design systems that allow agents to navigate a billion-line codebase with high precision.
  2. AI Workload Optimization: Ensure our AI infrastructure is performant and cost-effective as it scales to become our primary compute driver.
  3. Lead New Techniques: Develop new methods for synthetic data generation and agent evaluations that outperform current industry benchmarks.

Skills

Required

  • architecting reasoning layers
  • designing agent systems for large codebases
  • AI workload optimization
  • synthetic data generation
  • agent evaluation techniques
  • Reinforcement Learning from Compiler Feedback (RLCF)
  • Domain-Specific Distillation
  • fine-tuning models for code-specific constraints
  • continuous learning loops for models
  • system design for scale, correctness, and reliability

Nice to have

  • expertise in Luau language
  • understanding of proprietary high-performance engine architecture

What the JD emphasized

  • Reinforcement Learning from Compiler Feedback (RLCF)
  • Domain-Specific Distillation
  • synthetic data generation
  • agent evaluations

Other signals

  • building a code intelligence platform
  • architecting a reasoning layer for agents
  • handling mechanical toil of modern development
  • building HITL gates for AI-driven evolution
  • developing platforms and first-party agents