Machine Learning Engineer (Mid-Level)
Multiverse Computing2 days ago
Barcelona, SpainMid Level
Responsibilities
- Build and maintain multi-node training and post-training pipelines using SFT, distillation, and reinforcement learning.
- Develop scalable data pipelines for curation, filtering, deduplication, tokenization, and synthetic data generation.
- Build reproducible evaluation pipelines and reinforcement-learning environments, verifiers, and reward pipelines.
- Profile and optimize training and inference performance, including GPU utilization, memory, parallelism, checkpointing, and I/O.
- Package, optimize, and benchmark models for deployment on NVIDIA, non-NVIDIA, and edge hardware.
- Translate research ideas involving architectures, compression methods, and training recipes into clean, tested, reusable code.
- Maintain reproducibility through containers, configuration, experiment tracking, CI, code review, and documentation.
- Collaborate with research teams to run experiments and develop ML systems.
Requirements
- Master’s degree in Computer Science, AI, Mathematics, Physics, Engineering, or a related field, or equivalent industry experience.
- At least 2 years of industry experience building ML systems, including at least 1 year of hands-on transformer-model or LLM experience.
- Strong Python software engineering skills and experience with Git, code review, unit tests, integration tests, and CI.
- Strong PyTorch skills and hands-on experience training or fine-tuning transformer models.
- Experience with multi-GPU jobs and distributed training concepts including data, tensor, and pipeline parallelism, FSDP, or DeepSpeed.
- Working knowledge of Linux, containers such as Docker, and shared GPU clusters using Slurm or Kubernetes.
- Experience building data pipelines for large datasets and solid understanding of ML training fundamentals.
- Good debugging and communication skills and fluent English.
- Preferred qualifications include experience with NVIDIA ML tooling, post-training methods and frameworks, inference engines, edge or alternative hardware, model compression, GPU profiling, evaluation tooling, experiment tracking, MLOps, AWS, physical AI tooling, Spanish, open-source ML contributions, quantum-inspired methods, publications, or patents.
Benefits
- Indefinite contract, equal pay guarantee, variable performance bonus, and signing bonus.
- Work visa sponsorship if applicable and relocation package if applicable.
- Private health insurance and eligibility for an educational budget according to internal policy.
- Hybrid work opportunity and flexible working hours.
- Language classes and discounted lunch options.
- Career plan and opportunities to learn and teach.
- High-performance, collaborative environment working on cutting-edge technologies.
Categories