Machine Learning Engineer (Mid-Level)
Core
Build and maintain training, data, and evaluation pipelines for large language models (LLMs) and transformer architectures, optimizing performance for research and deployment.
Role type
Mid-level Machine Learning Engineer (LLM/Transformer focus)
Builds
Multi-node training pipelines, data curation/filtering pipelines, evaluation benchmarks, RL environments, and optimized model deployments.
Domain
Deep Tech / AI / Large Language Models / Quantum Software
Deliverable
production ML models
Required skills
Python, PyTorch, Transformer models, Distributed training (FSDP/DeepSpeed), Linux, Docker, Data pipelines, ML fundamentals
Preferred skills
Megatron-LM/NeMo, Post-training (SFT/DPO/RL), Inference engines (vLLM/TensorRT), Model compression, GPU profiling, MLOps, Spanish
Technologies
Megatron-LM, NeMo, PyTorch, Docker, Kubernetes, Slurm, vLLM, TensorRT-LLM, AWS, MLflow, Isaac Sim
Responsibilities
Build multi-node training and post-training pipelines; Build scalable data pipelines for curation and synthetic data; Build and maintain evaluation pipelines; Build RL environments and reward pipelines; Profile and optimize training/inference on GPU; Package and benchmark models for deployment; Turn research ideas into reusable code; Maintain reproducibility via containers and CI.
Seniority
Mid-level, hands-on IC
