People Operations AI Research Contributor
Aptura · Global
Infrastructureremote
At a glance
Highlights
- Create realistic headcount and compensation tasks to test frontier AI models.
- Define clear answers and yes-or-no rubrics to grade AI attempts.
- Turn real People Ops workflows into evaluations for AI training.
What they offer
- Immediate start for ~4-6 week project
- Task-based pay with weekly payments
- Independent contractor engagement with flexible schedule
Ideal candidate
Detail-oriented People Ops professional with hands-on modeling experience who has experimented with AI and cares about accuracy. Enjoys distilling real work into clear tasks and precise evaluation criteria.
Requirements
- 3+ years in people operations, HR analytics, total rewards or workforce planning
- Experience across headcount, compensation, benefits or org design
- Hands-on Excel building headcount and comp models, lookups, pivot tables and formatting
- Strong written English communication skills
- Detail-oriented with willingness to learn AI and evals concepts
- remote policy
- Fully remote, work on own schedule
- hiring process
- Apply with details and CV, join a 30-minute group walkthrough, then submit an original task brief with human-built input workbooks for frontier model testing and rubric review for approval.
Lab profile
About Aptura
Builds evaluation datasets and RL environments to make AI reliable in finance, healthcare, and legal.
What they build
- Evaluation datasets for regulated domains
- RL environments for high-stakes tasks
- Expert-curated training data and calibrated rubrics
Known for
- Research partner to Anthropic, OpenAI, Google DeepMind
- Specialization in finance, healthcare, and legal
- Clinician expert network for medical oversight
Key facts
- Based in London
- Small, lean team
Posted · Last seen