Staff AI Platform Engineer — ML & RL Lead (London, Hybrid)

Oliver Bernard; Paris; Hybride; EUR 222 000 - 257 000; Plein temps

Résumé du poste

Oliver Bernard is seeking a Staff / Principal AI Engineer to lead major areas across model improvement, RL, evaluations and agent infrastructure in a production AI platform. You will own research directions and stay close to models and experiments.

This is a deeply hands-on IC role with technical leadership, collaborating with teams, defining training-data pipelines, evaluation frameworks, and scalable systems. Location London, hybrid—2 days onsite per week.

Qualifications

• Strong hands-on experience training, fine-tuning or post-training ML models.
• Deep understanding of modern ML, RL, evaluation and model optimisation.
• Experience building agentic systems or AI infrastructure.
• Strong software engineering fundamentals and production-quality Python.
• Track record owning technically ambiguous, high-impact projects.
• Staff / Principal-level technical leadership while remaining an IC.

Responsabilités

• Own major AI initiatives from research question through to production.
• Lead work across post-training, fine-tuning, RL and model evaluation.
• Build agent harnesses, sandboxed execution environments and feedback systems.
• Develop training-data, synthetic-data and curation pipelines.
• Design experiments, evaluation frameworks and measurable success criteria.

Connaissances

Hands-on ML training

Fine-tuning models

Post-training ML

RL and evaluation

AI infrastructure

Python production-grade

Technical leadership IC

Ambiguity ownership

Description du poste

Oliver Bernard is seeking a Staff / Principal AI Engineer to lead major areas across model improvement, RL, evaluations and agent infrastructure in a production AI platform. You will own research directions and stay close to models and experiments.

This is a deeply hands-on IC role with technical leadership, collaborating with teams, defining training-data pipelines, evaluation frameworks, and scalable systems. Location London, hybrid—2 days onsite per week.