Prolific

Director of AI Research

Posted · Remote

Apply on Prolific

Director Of AI Research

Location:Based within commutable distance of at least one of our global hubs (London, New York or San Francisco).

Prolific

Prolific is not just another player in the AI space - we are the architects of the human data infrastructure that's reshaping the landscape of AI development. In a world where foundational AI technologies are increasingly commoditized, it's the quality and diversity of human-generated data that truly differentiates products and models.

The role

Our AI research team publishes at venues like ICML, ICLR and NeurIPS, works directly with frontier labs on their hardest data and measurement problems, and builds evaluations for public bodies shaping how AI is governed.

Further advances in frontier AI depend on better data and better measurement, and the science of both is still being written. Lead the team that pushes it forward. You'll run a team of research engineers and scientists who build evaluation methods, benchmarks and environments, and forward deployed researchers who work side by side with our closest partners. The team's work ships as artifacts the field requires to push the frontier: benchmarks, evaluations, environments and datasets that labs build on. Your job is to make Prolific a scientific leader in how human signal improves AI, how people and AI work together, and how AI accelerates research. Turn that work into methods, benchmarks and tools that customers use, and independent evidence the field can trust

You'll set the research direction with the team, partner with the product and engineering leaders who take our methods to market, and be Prolific's senior scientific voice with frontier labs, regulators and the research community, working directly with our CEO and executive team. It's a builder's role: a small, strong team, direct access to real, verified people from across the world's population, including leading domain experts, and room to define what good data and evaluation science looks like for the industry. You'll have the compute, budget and tooling to run large human studies and serious model experiments.


Problems You'll Work On

• Pushing model capability with expert data: which experts, which tasks, and how much signal each one adds.
• Agents and environments for training and evaluation, with oversight that scales.
• Safety and how AI affects people (persuasion, perception, reliance), measured with real, representative participants.
• Knowing what moves a model: measuring which data, experts and tasks actually drive capability gains, and why.
• Measurement you can trust: evals, validated judges and human baselines that hold up to saturation and contamination.
• Task and data design that moves a model: signal quality, difficulty and coverage.

What you’ll bring to the role

Research direction

  • Own a research strategy aimed at the goals we share with frontier labs: more capable models, safer models, and evidence people can trust. It spans work with lab partners, reusable benchmarks and evals, frontier bets, and public-benefit research.
  • Set the scientific bar for the team: rigorous methods and results that hold up to peer review and to scrutiny from frontier labs.
  • Make our research the source of what Prolific builds next: reusable methods, benchmarks, tools and new products.
  • Build a portfolio of public benchmarks, environments and datasets that become reference points for frontier labs.
  • Partner with the leads for each of our product areas, from agents and evaluation to multimodal, safety and specialist domains.
  • Protected time for novel and public-benefit research.

Partnerships and impact

  • Be trusted alone in the room with research leaders at frontier labs, and shape our most ambitious research collaborations with them.
  • Represent Prolific's science with regulators, safety institutes and academia. Build the partnerships and publications that make us credible.
  • Work with product and commercial leads to turn that portfolio into data and evaluations labs adopt and come back for.
  • Report research impact to the executive team and board in terms of customers, revenue and scientific standing.

Building the team

  • Lead, hire and grow a UK/US team of research engineers, research scientists and forward deployed researchers. Raise the bar with every hire.
  • Build a team where strong senior researchers lead their own agendas, and stay close enough to the work to shape the hardest problems yourself.

What you’ll be doing in the role

Must-haves

  • Research credibility in post-training data, RL environments, AI evaluation, or safety and human-AI interaction. A record frontier lab researchers recognise: papers, benchmarks or methods in use.
  • Strong grounding in experimental design and measurement, including with human participants: sampling, validity, reliability, bias.
  • 8+ years in research, including 3+ years leading researchers across locations (hiring, managing and growing senior people), ideally in industry or applied settings.
  • A track record of turning research into results customers use, including research artifacts others adopted (benchmarks, datasets, environments, tools), not only papers.
  • A clear point of view on where data and evaluation for frontier AI are going, the judgement to make hard portfolio calls, and the ability to make that case to executives, customers and the research community.
  • Hands-on enough to review a method, read the code and challenge a result.

Nice-to-haves

  • Experience at a frontier lab, an AI safety institute, or a human data or evals company.
  • A background that spans ML and behavioural or social science (psychology, psychometrics, economics, computational social science).
  • Experience building agentic evals or LLM-as-judge pipelines at scale.
  • Work with regulators or standards bodies on AI evaluation.

Why Prolific is a great place to work

We've built a unique platform that connects researchers and companies with a global pool of participants, enabling the collection of high-quality, ethically sourced human behavioral data and feedback. This data is the cornerstone of developing more accurate, nuanced, and aligned AI systems.

We believe that the next leap in AI capabilities won't come solely from scaling existing models, but from integrating diverse human

This listing is copied from the employer's official posting. Apply on the official page. We do not accept applications here.

Apply on Prolific