Staff Research Engineer - Post-Training & Sandbox Architect

Dex; Greater London; On-site; GBP 120,000 - 180,000; Full time

Job summary

Dex is partnering with frontier AI labs to identify opportunities in post-training, reinforcement learning, and environment design. The role of Member of Technical Staff sits between research and systems engineering, owning the full harness and data pipelines for model testing in complex digital tasks.

You will shape technical bets, curate large-scale training distributions, and build production-grade systems to test agents against multi-step tasks.

Qualifications

• Experience training/fine-tuning or post-training models (SFT/RL/RLHF).
• Strong foundations in ML and evaluation methodologies.
• Production-grade Python engineering with modular, observable systems.

Responsibilities

• Design and run end-to-end post-training pipelines (SFT/RL).
• Architect sandboxed execution environments and agent harnesses.
• Build scalable training-data infrastructure from messy data.
• Establish evaluation frameworks to benchmark capabilities.
• Provide technical leadership across ambiguous research directions.

Skills

SFT/RL

Python

ML fundamentals

Evaluation methods

System design

Tools

TensorFlow

PyTorch

Docker

Job description

Dex is partnering with frontier AI labs to identify opportunities in post-training, reinforcement learning, and environment design. The role of Member of Technical Staff sits between research and systems engineering, owning the full harness and data pipelines for model testing in complex digital tasks.

You will shape technical bets, curate large-scale training distributions, and build production-grade systems to test agents against multi-step tasks.