Senior Research Engineer/Scientist
ServiceNowMontrealFull TimeEngineering
Remotely
pythonllmtransformermegatrondeepspeeddpogrpoppo
Job Description
📋 Description
- Lead end-to-end AI/ML research for enterprise LLMs across data curation, pretraining, fine-tuning
- Build scalable training/evaluation methods and pipelines for reproducibility and measurable
- Collaborate with researchers, engineers, and product managers to deliver high-impact work.
- Own projects from design to delivery, translating requirements into results.
🎯 Requirements
- Post-training LLM expertise, including supervised fine-tuning and distillation.
- RL for LLMs experience (PPO, GRPO, DPO, reward modeling).
- Distributed/large-scale training with FSDP, DeepSpeed, Megatron; various parallelism strategies.
- Synthetic data generation and large-scale data curation including deduplication and filtering.
- Experience with transformer architectures (decoder-only, encoder-decoder, MoE).
- Advanced Python, strong OOP and design-pattern skills; rapid prototyping.