Jobs
My ads
My job alerts
Sign in
Find a job Career Tips Companies
Find

Research engineer (llm training and performance)

London
Permanent
JetBrains
Research engineer
€75,000 a year
Posted: 27 December
Offer description

Research Engineer (LLM Training and Performance)

At JetBrains, code is our passion. Ever since we started back in 2000, we have been striving to make the strongest, most effective developer tools on earth. By automating routine checks and corrections, our tools speed up production, freeing developers to grow, discover, and create. We’re looking for a Research Engineer who will own the training stack and model architecture for our Mellum LLM family.


Responsibilities

* Be responsible for improving end-to-end performance for multi-node LLM pre-training and post-training pipelines.
* Profile hotspots (Nsight Systems/Compute, NVTX) and fix them using compute/comm overlap, kernel fusion, scheduling, etc.
* Design and evaluate architecture choices (depth/width, attention variants including GQA/MQA/MLA/Flash-style, RoPE scaling/NTK, and MoE routing and load-balancing).
* Implement custom ops (Triton and/or CUDA C++), integrate via PyTorch extensions, and upstream when possible.
* Push memory/perf levers: FSDP/ZeRO, activation checkpointing, FP8/TE, tensor/pipeline/sequence/expert parallelism, NCCL tuning.
* Harden large runs by building elastic and fault-tolerant training setups, ensuring robust checkpointing, strengthening reproducibility, and improving resilience to preemption.
* Keep the data path fast using streaming and sharded data loaders and tokenizer pipelines, as well as improve overall throughput and cache efficiency.
* Define the right metrics, build dashboards, and deliver steady improvements.
* Run both pre-training and post-training (including SFT, RLHF, and GRPO-style methods) efficiently across sizable clusters.


Qualifications

* Strong PyTorch and PyTorch Distributed experience, having run multi-node jobs with tens to hundreds of GPUs.
* Hands‑on experience with Megatron‑LM/Megatron‑Core/NeMo, DeepSpeed, or serious FSDP/ZeRO expertise.
* Real profiling expertise (Nsight Systems/Compute, nvprof) and experience with NVTX‑instrumented workflows.
* GPU programming skills with Triton and/or CUDA, and the ability to write, test, and debug kernels.
* A solid understanding of NCCL collectives, as well as topology and fabric effects (IB/RoCE), and how they show up in traces.


Ideal Candidate Experience

* FlashAttention‑2 and 3, CUTLASS and CuTe, TransformerEngine and FP8, Inductor, AOTAutograd, and torch.compile.
* MoE at scale (expert parallel, router losses, capacity management) and long‑context tricks (ALiBi/YaRN/NTK scaling).
* Kubernetes or SLURM at scale, placement and affinity tuning, as well as AWS, GCP, and Azure GPU fleets.
* Web‑scale data plumbing (streaming datasets, Parquet and TFRecord, tokenizer perf), eval harnesses, and benchmarking.
* Safety and post‑training methods, such as DPO, ORPO, GRPO, and reward models.
* Inference ecosystems such as vLLM and paged KV.

We process the data provided in your job application in accordance with the Recruitment Privacy Policy.

#J-18808-Ljbffr

Apply
Create E-mail Alert
Job alert activated
Saved
Save
Similar job
Machine learning research engineer - speech/audio/gen-ai - 6 month fixed term contract
Staines
Permanent
Temporary
SAMSUNG
Research engineer
€55,000 a year
Similar job
Ml research engineer
London
Permanent
Symbolica
Research engineer
€80,000 a year
Similar job
Research engineer, multimodal and video modeling
London
Permanent
The Rundown AI, Inc.
Research engineer
€70,000 a year
See more jobs
Similar jobs
Engineering jobs in London
jobs London
jobs Greater London
jobs England
Home > Jobs > Engineering jobs > Research engineer jobs > Research engineer jobs in London > Research Engineer (LLM Training and Performance)

About Jobijoba

  • Career Advice
  • Company Reviews

Search for jobs

  • Jobs by Job Title
  • Jobs by Industry
  • Jobs by Company
  • Jobs by Location
  • Jobs by Keywords

Contact / Partnership

  • Contact
  • Publish your job offers on Jobijoba

Legal notice - Terms of Service - Privacy Policy - Manage my cookies - Accessibility: Not compliant

© 2026 Jobijoba - All Rights Reserved

Apply
Create E-mail Alert
Job alert activated
Saved
Save