Frontier Data Strategist

Scale AI · San Francisco, CA; New York, NY · Gen AI Operations · listed April 20, 2026

The shape of it

Seniority
Not stated
Where
Not stated
Stated pay
$160,800 – $241,000 USD
Requirements listed
6
Length
1,076 words

In the posting’s own words

Frontier AI labs decide what direction their models need to take before the actual training starts. As a Frontier Data Strategist on Scale's Generative AI team, you'll be a trusted voice in those decisions. You'll partner with research leads and senior stakeholders at the world's leading AI labs to shape the data strategy behind their next generation of models, then design the human-data programs to deliver it.

What it asks for · 6

  • Deep fluency in the LLM training lifecycle, especially post-training: SFT, preference data, reward modeling, RL environments, evaluations, agentic training, etc.
  • A track record of owning relationships with senior technical stakeholders, and of influencing their direction through substance.
  • Experience influencing a product or research roadmap without direct authority.
  • The ability to read a research paper and form an opinion on what data would improve a given capability, and to defend that opinion to researchers.
  • A degree in CS, engineering, math, physics, or another quantitative field.
  • A graduate degree (PhD, MS) or published research in ML or a related field.

What the job covers

  • Own relationships with senior research stakeholders at your assigned frontier labs, such as research leads, heads of post-training, and technical leadership, and become a strategic thought partner on their post-training, evaluation, and agent roadmaps.
  • Set the data strategy for your accounts: anticipate needs across research areas, spot where the biggest opportunities are, and build a multi-program plan for how Scale can help.
  • Direct the cross-functional team designing Scale's most complex and novel human-data programs, including work that has no precedent at Scale.
  • Shape Scale's product roadmap by representing your customers in planning with AI Product Managers and ML Research, and by resolving tradeoffs between customer needs and what Scale can build.
  • Bring insights from the frontier (research directions, emerging data needs, capability gaps) back to Scale's leadership and technical teams to guide strategy and investment decisions.
  • Own the pipeline of new opportunities for your accounts. Lead scoping, forecasting, and pricing with Delivery, Growth Operations, and Finance, and see the work through to signed programs and a clean handoff.
  • Mentor Frontier Data Strategists, set the bar for program design and customer engagement, and help build the playbooks the team runs on.

Degree language

  • A degree in CS, engineering, math, physics, or another quantitative field.
  • A graduate degree (PhD, MS) or published research in ML or a related field.

Tools and skills named

Models & research
  • Machine learning5×
  • Evaluations2×
  • LLM
Product & design
  • Roadmap3×
  • Product management
Go to market
  • Forecasting2×
  • Customer success
Ways of working
  • Cross-functional2×
  • Mentorship

Words the posting leans on

  • data14×
  • research13×
  • frontier9×
  • technical8×
  • product7×
  • programs7×
  • customer6×
  • labs5×
  • senior5×
  • experience4×
  • frontier data4×
  • leading4×
  • post-training4×
  • roadmap4×
  • training4×
  • build3×

Counted from the posting after the mission statement and the legal notices are set aside. The ones near the top are the ones a screener is looking for.

The posting, your resume, and the gaps between them. One click loads all three.

More open at Scale AI

every open role at Scale AI →

How this page was made

An automated read of a public job posting, fetched November 2, 2026 and last changed by Scale AI on October 28, 2026. Every list above is pulled from the posting’s own sentences — nothing rewritten, nothing added, no judgment about the role or the company. Counts and seniority are read off the text by rule, so they can be wrong where the posting is unusual. The original is the only thing that binds. Openings close without warning; check the source before spending an evening on it.