Forward Deployed Researcher - Data as a Service

Snorkel AI Snorkel AI · Data AI · Redwood City, CA +1 · Remote · 415 - DaaS Sales & Success

This role focuses on partnering with frontier AI research labs to design datasets and environments that improve model performance. It involves scoping training data needs, designing RL environments, developing evaluation frameworks, probing model behavior, and translating research objectives into technical plans. The role is customer-facing and requires strong technical and research credibility.

What you'd actually do

  1. Partner with frontier AI research labs to design datasets and environments that improve model performance
  2. Lead technical conversations with customer researchers to understand model capabilities, failure modes, data requirements, and success criteria
  3. Probe model behavior through systematic evaluation to uncover weaknesses and identify high-impact data interventions
  4. Design evaluation frameworks, calibration processes, and quality rubrics that establish measurable project success metrics
  5. Develop technical specifications for data projects that balance research rigor with operational feasibility

Skills

Required

  • Python
  • ML frameworks
  • LLM APIs
  • communication skills
  • technical presentations

Nice to have

  • frontier AI concepts
  • LLMs
  • training data pipelines
  • evaluation methodologies
  • post-training techniques (RLHF, DPO, RLAIF)
  • coding agents
  • reasoning
  • multimodal models
  • RL environments
  • applied ML research
  • data science
  • customer-facing or collaborative research experience
  • data curation workflows
  • synthetic data generation
  • LLM-as-a-Judge
  • evaluation framework design
  • fast-moving environment
  • ambiguity
  • rapid iteration
  • B.S. in Computer Science, Machine Learning, or related field with 4+ years of experience in AI/ML research or technical roles
  • advanced degree preferred

What the JD emphasized

  • customer-facing
  • frontier AI research labs
  • model performance
  • datasets and environments
  • evaluation frameworks
  • model behavior
  • failure modes
  • data requirements
  • training data needs
  • RL environments
  • post-training techniques
  • evaluation methodologies

Other signals

  • customer-facing
  • research
  • data
  • evaluation
  • model performance