AI Research Engineer - Distributed Training / LLMs

European Tech Recruit; Paris; Sur place; EUR 90 000 - 130 000; Plein temps

Résumé du poste

A cutting-edge AI safety startup is seeking an AI Research Engineer in Paris to tackle complex training challenges for LLMs and VLMs. The role focuses on designing next-generation architectures, building multimodal training pipelines, and optimizing performance through advanced distributed training frameworks like Megatron and DeepSpeed. Candidates should possess a deep understanding of distributed training concepts and Triton kernels. This is a full-time position with a mid-senior level requirement, offering the opportunity to drive innovation in advanced AI systems.

Qualifications

• Deep understanding of tensor parallelism, pipeline parallelism, expert parallelism, ZeRO, and FSDP.
• Ability to write Triton kernels for custom operators or fundamental layers like attention and MLP.
• Experience working with multimodal systems like Llava-style and Flamingo-style.

Responsabilités

• Train LLMs and VLMs using advanced distributed training frameworks.
• Design and implement next-generation architectures, including Mixture-of-Experts (MoE).
• Build and maintain complex multimodal training pipelines for text, images, and audio data.
• Develop custom Triton kernels to optimize training and inference bottlenecks.
• Drive innovation exploring new model architectures and hyperparameter strategies.
• Focus on maximizing hardware throughput and ensuring training stability and convergence.

Connaissances

Distributed training concepts

Triton kernels

Multimodal systems

ZeRO

FSDP

Triton kernels

Multimodal systems

Distributed training

Outils

Megatron

DeepSpeed

Triton

Description du poste

We are partnered with a cutting‑edge AI safety startup that is building the foundational reliability and optimization layer for advanced AI systems capable of managing hundreds of millions of API calls monthly and training proprietary LLMs and VLMs that outperform commercial and open‑source models. They are seeking a highly focused AI Research Engineer to tackle complex, production‑critical training challenges.

This is a permanent opportunity based in Paris, France.

Key Responsibilities for this AI Research Engineer position:

• Train LLMs and VLMs using advanced distributed training frameworks (e.g., Megatron, DeepSpeed).
• Design and implement next‑generation architectures, including Mixture‑of‑Experts (MoE), to achieve high‑efficiency scaling.
• Build and maintain complex multimodal training pipelines that effectively handle text, images, and audio data.
• Develop custom Triton kernels to identify and optimize training and inference bottlenecks.
• Drive innovation by exploring new model architectures, hyperparameter strategies, and dataset compositions.
• Focus intensely on maximizing hardware throughput and ensuring training stability and convergence.

Key Requirements:

• Deep understanding of distributed training concepts: tensor parallelism, pipeline parallelism, expert parallelism, ZeRO, and FSDP.
• Ability to write Triton kernels to accelerate custom operators or fundamental layers like attention and MLP.
• Experience working with multimodal systems (e.g., Llava‑style, Flamingo‑style).

Keywords: AI Research Engineer / LLM / Large Language Models / VLM / Vision‑Language Models / Distributed Training / Megatron / DeepSpeed / Triton Kernels / Multimodal AI / Mixture‑of‑Experts / MoE / FSDP / ZeRO / AI Safety / Paris / Optimization / Hardware Throughput

Seniority level: Mid‑Senior level; Employment type: Full‑time; Job function: Engineering and Design; Industries: Software Development and Research Services.

If you are interested in this AI Research Engineer position, please send a copy of your CV to [e-mail]

By applying to this role you understand that we may collect your personal data and store and process it on our systems. For more information please see our Privacy Notice https://eu-recruit.com/wp-content/uploads/2024/07/European-Tech-Recruit-Privacy-Notice-2024.pdf