Hybrid ML Research Engineer - LLM Training & Safety

White Circle; Paris; Hybrid; EUR 90,000 - 130,000; Full time

Benefits offered by this job

Paid time off

Hybrid work (Paris/London)

Relocation support for Paris

Medical insurance for France-based

Hardware, tools and services

Team off-sites

Job summary

White Circle is seeking several ML Engineers to train, post-train, and evaluate the LLMs at the core of our platform. You will work hands-on with large-scale data pipelines, SFT/RLHF-style alignment, and distributed multi-GPU training to push safety and reliability.

Join a focused team building scalable representations, labeling, and diagnostics that feed product decisions and operations. Hybrid Paris/London roles with relocation support for Paris.

Qualifications

• Strong Python + SQL with an engineering mindset.
• Solid applied NLP/ML experience on real-world text.
• Experience with large-scale data processing and pipelines.
• Familiarity with safety/moderation datasets or policy/rule systems is a plus.
• Multilingual model training experience is a plus.

Responsibilities

• Turn petabytes of unstructured text into a structured, explorable view.
• Build scalable representation pipelines: sampling, preprocessing, embeddings, indexing, retrieval.
• Use LLMs pragmatically: labeling, weak supervision, data enrichment, diagnostics.
• Deliver insights that translate findings into product/actions.
• Ship self-serve analytics: datasets, data models, dashboards.
• Partner with engineering/research to align pipelines with production constraints.

Skills

Python

SQL

NLP/ML

Distributed training

Data pipelines

Model evaluation

Tools

PyTorch

HuggingFace

TensorFlow

Job description

White Circle is seeking several ML Engineers to train, post-train, and evaluate the LLMs at the core of our platform. You will work hands-on with large-scale data pipelines, SFT/RLHF-style alignment, and distributed multi-GPU training to push safety and reliability.

Join a focused team building scalable representations, labeling, and diagnostics that feed product decisions and operations. Hybrid Paris/London roles with relocation support for Paris.