About the job
About Aptura
Aptura specializes in creating evaluation datasets and reinforcement learning (RL) environments that enhance the reliability of AI across high-stakes sectors such as finance, healthcare, and legal domains. Our dedicated team curates expert training data, develops calibrated evaluation rubrics, and constructs verifiable task environments tailored for AI labs and startups that are innovating within regulated industries.
As a compact and agile team based in London, we prioritize speed and take our work seriously. Each team member plays a crucial role, where initiative is celebrated, and ownership is a fundamental expectation. If you aspire to influence how advanced AI systems learn to function in real-world applications, we encourage you to apply.
About the Role
In the role of a Technical Staff Member on our Applied AI team, you will create the tasks and environments essential for AI labs to train and assess their agents across finance, healthcare, and legal sectors.
Your daily responsibilities will include constructing RL environments using spreadsheets, documents, and professional workflows. You will develop verification logic and reward functions, collaborating closely with domain experts to define the criteria for accurate responses in complex models such as leveraged buyouts or clinical notes. Your work may oscillate between engineering and research, but the unifying objective is to deliver the ground truth against which leading models are evaluated.
What You'll Do
- Develop RL environments tailored to the finance, healthcare, and legal sectors.
- Collaborate in creating tasks with definitive answers, calibrated evaluation criteria, and automated reward signals.
- Design verification logic and reward functions to differentiate between effective and ineffective model outputs.
- Engage with domain experts (investment analysts, healthcare professionals, attorneys) to convert intricate professional workflows into structured tasks.
- Prototype innovative methods for evaluation, verification, and synthetic data generation.
Who We're Looking For
- Hands-on experience with large language models (LLMs): including prompting, evaluation, and agentic harnesses. You have developed functional systems instead of merely theoretical models.
- High initiative and technical proficiency. You don’t wait for permission, specifications, or a detailed plan. You identify what needs to be accomplished, determine how to achieve it, and execute effectively.
- Ability to adapt to varied contexts. Your role will involve switching between engineering tasks, evaluation design, and close collaboration with domain experts frequently within a single day.
- You are proactive in shipping and iterating. In our small team, there’s no room for work that remains in limbo.
