Researcher - Reinforcement Learning

Huawei Technologies Co.; Edmonton; Hybrid; CAD 120,000 - 180,000; Full time

Job summary

Huawei Canada is hiring a Reinforcement Learning Researcher on a 12-month contract to advance agentic learning for LLMs and tool-use agents. The role focuses on post-training RL paradigms, evaluation benchmarks, and scaling experiments on large GPU clusters.

The ideal candidate has a PhD or equivalent, strong deep learning background with Transformers, and experience in RL, self-supervised learning, and language model fine-tuning. Excellent communication skills are required.

Qualifications

• PhD in CS or related field or master's with comparable experience.
• Strong foundation in deep learning with Transformers and optimization for large models.
• Practical or research experience in reinforcement learning, self-supervised learning, or language model fine-tuning.
• First-author paper in top venues (NeurIPS, ICML, ICLR, CVPR, ICCV, ECCV, ICRA).
• Proficiency in Python and distributed training tools (PyTorch/DeepSpeed/Megatron).
• Familiarity with post-training LLM pipelines (RLHF, PPO, SFT, LoRA, MoE).
• Experience with multi-agent RL, tool-use and browser/coding agents.
• Strong written and verbal communication; open research mindset.

Responsibilities

• Implement and evaluate training and evaluation pipelines for reasoning-enhanced LLMs and tool-using agents.
• Scale experiments on large GPU clusters and report scientific insights.
• Contribute to publications and collaborative research efforts.

Skills

Deep learning

Transformers

Reinforcement learning

Self-supervised learning

Language model fine-tuning

Python

PyTorch

DeepSpeed

Megatron

Distributed training frameworks

RLHF / PPO / SFT / LoRA / MoE

Multi-agent RL

Tool-use / browsing agents

Strong communication

Education

PhD in Computer Science or related field

Master's with comparable experience

Tools

PyTorch

DeepSpeed

Megatron

Job description

Huawei Canada has an immediate 12-month contract opening for a Reinforcement Learning Researcher.

About the team:

Founded in 2012, the Noah’s Ark lab has evolved into a prominent research organization with notable achievements in academia and industry. The lab’s mission focuses on advancing artificial intelligence and related fields to benefit the company and society. Driven by impactful, long-term projects, the aim is to enhance state-of-the-art research while integrating innovations into the company's products and services, including LLMs, RL, NLP, computer vision, AI theory, and Autonomous driving.

About the job:

• Enabling Large Language Models (LLMs) to learn from experience, interaction, and environment feedback, moving beyond static fine-tuning toward continual, agentic self-improvement.




• LLM post-training paradigms (e.g., RLHF, GRPO, reward-free methods, etc.).




• Agentic reinforcement learning for tool-using and browsing-based LLMs trained in interactive environments.




• Agentic evaluation and benchmarking, including design of multi-turn, verifiable reasoning tasks.




• Your work will involve implementing and evaluating new training and evaluation pipelines for reasoning-enhanced LLMs and tool-using agents, scaling experiments on large GPU clusters, and contributing to scientific insights and publications in this emerging area.

About the ideal candidate:

• PhD degree in Computer Science or related fields or master's degree with comparable experience.




• Strong foundation in deep learning, including architectures such as Transformers and optimization techniques for large models.




• Practical or research experience in reinforcement learning, self-supervised learning, or language model fine-tuning.




• Proven research record in AI by having at least one paper as the first author in top tier venues, such as NeurIPS, ICML, ICLR, CVPR, ICCV, ECCV, ICRA.




• Solid proficiency in Python and experience with PyTorch, DeepSpeed, Megatron and other distributed training frameworks.




• Familiarity with LLM post-training pipelines (RLHF, GRPO/PPO, SFT, LoRA, MoE, etc.) is an asset.




• Experience with multi-agent RL, tool-use / browser/coding agents, is an asset.




• Strong communication and writing skills; enthusiasm for open research and collaborative problem-solving.

Informations supplémentaires :

Huawei Canada s'engage à un processus de recrutement équitable, inclusif et accessible. Si vous avez besoin d'un accommodement à n'importe quelle étape du processus d'embauche, veuillez nous en informer, et nous travaillerons avec vous pour répondre à vos besoins.

Toutes les candidatures pour ce poste sont examinées directement par notre équipe de recrutement, nous n'utilisons pas d'outils d'intelligence artificielle pour filtrer ou sélectionner les candidats.

Huawei vise à soutenir un environnement de travail francophone pour ses employés au Québec. Nous avons pris des mesures pour éviter de demander une autre langue que le français pour ce poste. Cependant, la maîtrise de l'anglais est essentielle pour ce rôle pour les raisons suivantes :

La personne devra communiquer régulièrement avec des collègues situés en dehors du Québec, où l'anglais est la langue principale utilisée pour la communication entre les bureaux. De plus, la nature des tâches liées à ce poste, qui relève d'un domaine hautement spécialisé de l'intelligence artificielle, nécessite également une connaissance de l'anglais.

L'utilisation du genre masculin a été adoptée afin de faciliter la lecture et n'a aucune intention discriminatoire.