Member of Technical Staff - AI Engineer

This role is with one of Dex's trusted partner companies. We work closely with their teams to truly understand their culture, goals, and what they're looking for, so we can match you with the right opportunity and give you context about the role before you commit to a process. If you're interested sign up to Dex to apply. Dex is an AI recruiter agent that helps you run your job search. Tell Dex your stack, seniority, and what you want to build. We will manage your applications and surface other opportunities that are a fit. The role Frontier AI systems have hit the limits of static web scrapes and now require active, trial-and-error experience on real software to improve. Operating out of engineering hubs in London and Paris and collaborating closely with the world's leading AI labs, this team builds resettable digital environments where autonomous agents navigate real-world software, fail, learn, and generate the experiential datasets powering the next generation of foundation models. You will join as a Member of Technical Staff owning end-to-end research engineering across model post-training, evaluation loops, and sandbox execution systems. This is a hands-on systems and model engineering seat, not a prompt-wrapper role and not an ivory-tower research track. You will turn open-ended questions into working technical architectures, run training and fine-tuning pipelines, and ship production-grade infrastructure that directly expands what frontier models can do. The work

• Fine-tune and post-train models using SFT, reinforcement learning, and targeted feedback loops, moving far past public API integration into direct model adaptation.
• Build and scale data creation and curation pipelines that extract high-signal synthetic and experiential training data from agent task trajectories.
• Design evaluation harnesses and automated benchmark environments that measure agent performance, task completion, and regression across complex digital tasks.
• Develop resilient agent harnesses and sandboxed execution environments that reliably interface models with real desktop, web, and enterprise applications.
• Prototype novel training and agent workflows rapidly, isolate what holds up against real task distributions, and productionise winning approaches alongside software engineers. What You Bring
• Direct, hands-on experience training, fine-tuning, or post-training models with methods like SFT, RL, or RLHF, with deep knowledge of what changed and why.
• Strong software engineering fundamentals in Python, producing clean, tested, and observable production-quality systems rather than disposable research scripts.
• Proven track record of breaking down ambiguous AI problems from first principles and driving substantial technical initiatives across multi-month horizons.
• Demonstrated technical leadership without formal management authority, setting technical direction and coordinating complex projects across peer engineers.
• Ability to work on-site from engineering hubs in London or Paris. Why apply through Dex Frontier AI engineering roles are heavily oversubscribed, and it is exceptionally difficult to separate genuine post-training work from standard wrapper applications through a conventional resume screen. Applying through Dex bypasses the initial recruiter noise, ensuring your specific model-training depth and engineering trade-offs are evaluated directly by the technical team. You sign up once, receive an unvarnished technical briefing on the team and architecture before speaking with them, and get matched to similar high-signal AI seats automatically. If you're interested, sign up to Dex to apply - https://jobs.meetdex.ai/jobs/aaf918d0-9b89-43dc-8e50-c864660f336d As part of the recruitment process at Dex, we process your personal data in accordance with our Privacy Notice for Job Applicants. This notice explains how and why your data is collected and used, and how you can contact us if you have any concerns.