Lead Architect — Large-Scale RL for Agentic LLMs

Thomson Reuters

Zug

On-site

CHF 150,000 - 210,000

Full time

2 days ago
Be an early applicant

A complete application in a minute

Tailored resume and cover letter, ready to send.

Get past ATS filters

Benefits offered by this job

Hybrid Work Model
Career Development
Mental Health Days / Wellbeing

Job summary

Thomson Reuters Labs is seeking a Lead Research Engineer to architect and optimize the infrastructure for training agentic large language models (LLMs) with reinforcement learning (RL). You will lead the RL training stack, optimize throughput, and ensure end-to-end reliability across training and inference at frontier scale.

You will collaborate with researchers to rapidly test new algorithms and environments, mentor engineers, and set engineering standards to enable scalable, reproducible

Qualifications

  • MSc or PhD in computer science, machine learning or related field, or equivalent industry experience.
  • 3+ years of professional experience in ML engineering or research engineering on large-scale model training.
  • Extensive experience training large models on multi-node GPU clusters with PyTorch and Megatron-LM/Megatron-Core, FSDP/DTensor or DeepSpeed.
  • Deep understanding of parallelism strategies (tensor, pipeline, context and expert parallelism) and their memory and communication trade-offs, esp. for MoE and long-context models.
  • Hands-on experience with LLM inference engines (vLLM, SGLang), incl. serving large MoE models.
  • Strong skills in GPU profiling and optimisation (memory, NCCL, CUDA graphs).
  • Strong software engineering practice and ability to work inside large open-source codebases.

Responsibilities

  • Lead technical direction: Own architecture of large-scale RL training stack.
  • Optimise full RL loop: generation, tool calls, reward computation, training and weight sync.
  • Build asynchronous, fault-tolerant pipelines where models interact with tools during training.
  • Debug numerical, distributed and performance issues across training, inference and model code.
  • Enable research by enabling quick testing of new algorithms and environments at scale.
  • Mentor and raise the bar: set standards for code quality, testing and reproducibility.

Skills

3+ years ML engineering
Technical leadership
Root-cause problem solving
Engineering rigour
Collaboration

Education

MSc or PhD in computer science or ML

Tools

PyTorch
Megatron-Core
FSDP/DTensor
DeepSpeed
vLLM / SGLang

Job description

Thomson Reuters Labs is seeking a Lead Research Engineer to architect and optimize the infrastructure for training agentic large language models (LLMs) with reinforcement learning (RL). You will lead the RL training stack, optimize throughput, and ensure end-to-end reliability across training and inference at frontier scale.

You will collaborate with researchers to rapidly test new algorithms and environments, mentor engineers, and set engineering standards to enable scalable, reproducible

Get your free, confidential resume review.

or drag and drop your file here.

Similar jobs

Similar jobs worth comparing

Lead RL Architect for Frontier-Scale Agentic LLMs
Lead RL Architect for Frontier-Scale Agentic LLMs

Thomson Reuters Corp. • Basel

Hybrid
CHF 190,000 - 270,000
Hybrid Work Model
Flexible work arrangements
Career development and growth
Lead Agentic RL Scientist for Grounded LLMs
Lead Agentic RL Scientist for Grounded LLMs

Thomson Reuters Corp. • Basel

Hybrid
CHF 130,000 - 190,000
Hybrid Work Model
Flex My Way
Career Development
+3
Lead Research Scientist: Agentic RL for LLMs
Lead Research Scientist: Agentic RL for LLMs

Thomson Reuters • Zug

On-site
CHF 180,000 - 240,000
Lead Agentic RL Scientist: Grounded LLMs & Tool Use
Lead Agentic RL Scientist: Grounded LLMs & Tool Use

Thomson Reuters • Zug

Hybrid
CHF 150,000 - 190,000
Hybrid work model
Flexible working arrangements
Career development programs
+2
Lead Research Scientist, Agentic RL for LLMs
Lead Research Scientist, Agentic RL for LLMs

Thomson Reuters Corp. • Zug

On-site
CHF 180,000 - 240,000
Hybrid work model
Flexible vacation
Mental health days
+3
Lead ML Platform Engineer – Hybrid LLM Infra
Lead ML Platform Engineer – Hybrid LLM Infra

Thomson Reuters Corp. • Basel

Hybrid
CHF 150,000 - 210,000
Hybrid work environment
Career development and growth programs
Mental health days and wellness
Lead ML Platform Engineer: Scale Trusted LLM Infrastructure
Lead ML Platform Engineer: Scale Trusted LLM Infrastructure

Thomson Reuters • Zug

Hybrid
CHF 180,000 - 250,000
Hybrid work model
Flexible work policy
Career development & growth
+4
Lead AI Engineer, Agent Platform for Legal Tech
Lead AI Engineer, Agent Platform for Legal Tech

CAN002 Thomson Reuters Canada Ltd • Zug

Hybrid
CHF 148,000 - 198,000
Hybrid Work Model
Flex My Way
Grow My Way
+5
Lead Research Engineer, Large-Scale Agentic RL Training
Lead Research Engineer, Large-Scale Agentic RL Training

Thomson Reuters Corp. • Basel

Hybrid
CHF 190,000 - 270,000
Hybrid Work Model
Flexible work arrangements
Career development and growth
Frontier AI Research Leader — Build & Grow Impactful Team
Frontier AI Research Leader — Build & Grow Impactful Team

Thomson Reuters • Zug

Hybrid
CHF 180,000 - 280,000
Hybrid Work Model
Flex My Way policy
Grow My Way skills development
+3
Top locations