R-001 – Research Engineer, LLM in Paris

Full time, Permanent position Research Paris, France Published on 17 September 2026

About Us

At OVHai LLM, we build large language models and design platforms and services around these models for specific use cases.

We’re a tight-knit team of 15 researchers and engineers, united by curiosity and a shared drive to push the boundaries of what LLMs can do — from training and fine-tuning to rigorous evaluation and benchmarking.

Our team is genuinely diverse and multicultural, bringing together different backgrounds, perspectives, and ways of thinking.

That mix is core to how we work: we believe the best research comes from teams that challenge each other’s assumptions and learn from one another every day.

We take our work seriously, but not ourselves.

Expect a collaborative, low-ego environment, plenty of interesting debates, shared lunches, and lots of fun — alongside real opportunities to contribute to cutting-edge LLM research from day one.

The role

This is a junior-level position for someone who has just finished their studies (university or master’s degree in computer science) and wants to work on training frontier LLMs.

You will work directly with senior engineers and are expected to have a solid understanding of deep learning concepts and some hands‑on experience with training small‑scale LLMs.

You might work on one or several stages of the pipeline, which includes data, training, post‑training and evaluation.

Your exact focus will be determined during the interview process, based on your background and the team’s needs.

What you will do

• Prepare and analyse training data: write filtering and deduplication scripts, inspect samples, build small quality classifiers, generate and validate synthetic data.
• Fine‑tune models (SFT, DPO, RL) on domain data and measure the effect.
• Run and extend our evaluation suite: add benchmarks, build small domain‑specific test sets, analyse failures.
• Help operate training runs on the cluster: launch jobs, monitor metrics, spot problems, write the tools to fix the issues.
• Reproduce results from papers on small models and report what does and does not transfer.
• Contribute to internal tools, dashboards and documentation.
• What we look for
• Master’s degree (or equivalent) in computer science, machine learning, applied mathematics or a related field; Ph D students finishing up are welcome.
• Solid Python. You write code other people can read, and you know how to use git, write tests.
• Hands‑on experience with Py Torch and the Hugging Face ecosystem (transformers, datasets, tokenizers, trl or peft) through coursework, internships or personal projects.

You have fine‑tuned at least one model yourself and can explain what happened.

• Good understanding of how transformers work (attention, positional encodings, tokenisation, training objectives) and of the basic LLM training lifecycle (pretraining, SFT, RLHF/DPO, RLVR).
• Comfortable with Linux, the command line and working on remote machines.
• Curiosity, rigour when looking at data, and the ability to say “I don’t know, let me check”.
• Fluent in both English and French (written and spoken).
• Nice to have
• An internship or thesis in NLP, LLMs or ML systems.
• Having launched a job on a GPU cluster (Slurm) or used multi‑GPU training even at small scale.
• Familiarity with evaluation tooling (lm‑eval‑harness) or with v LLM for fast inference.
• Some exposure to RL (a course, a project, an RLHF/GRPO notebook).
• Experience with data processing at scale (Spark, Ray, datatrove, Duck DB, large‑file handling).
• Interest in efficient architectures, long context, or non‑Transformer models.
• A Git Hub, a blog post or a Kaggle / Hugging Face profile that shows what you like to build.
• The Recruitment Process
• A 15-minute introductory call with a senior researcher to discuss the role
• An interview and technical use case assessment with two members of our research team
• An HR interview to assess soft skills
• Debriefing session
• #J-18808-Ljbffr