Principal Machine Learning Engineer

$137,600 - $267,000/Yr

Microsoft - Cambridge, MA

Full-time - Principal

Cambridge, MA

Publishing Industries

About the position

The Principal Machine Learning Engineer will play a crucial role in developing industry-leading machine learning solutions at Microsoft. This position requires a blend of analytical and development skills, with a focus on designing, training, and tuning machine learning models and their supporting infrastructure. The role involves collaborating with researchers and product teams to innovate and implement novel ML algorithms, while applying sound engineering principles to push the boundaries of current conventions.

Responsibilities

• Collaborate with a cross-disciplinary team of engineers, data scientists, and researchers in a research-oriented environment.
• Partner with researchers to learn, implement, and evolve novel ML algorithms and approaches.
• Interact with Microsoft product teams, MSR partners, and open-source communities to disseminate research implementations.
• Engineer complex, high-performance ML training jobs using advanced tools and frameworks.
• Simplify ML infrastructure complexities across research projects, including onboarding and debugging issues, while maintaining high GPU utilization.

Requirements

• Bachelor's Degree in Computer Science, Engineering, Statistics, Mathematics, or related field AND 6+ years of technical engineering experience with coding in languages such as Python, C, C++, C#, Java, or equivalent experience.
• 4+ years of industry experience in applying, implementing, and/or developing algorithms for machine learning, artificial intelligence, or statistics.
• 2+ years of industry experience with common ML engineering programming languages and platforms such as Python, PyTorch, TensorFlow, and/or HuggingFace.
• 2+ years of industry experience with building, debugging, and optimizing large, high-performance ML training jobs.

Nice-to-haves

• Masters or Doctorate in Computer Science, Machine Learning, Statistics, or equivalent technical field.
• Deep understanding of deep learning model architectures.
• Experience with low-level, high-performance ML frameworks such as CUDA, Triton, DirectML, or similar frameworks.
• Experience with large language models (LLM) architectures, inference, and fine-tuning.
• Experience with ML infrastructure including job schedulers, big data storage, low latency interconnects, and large clusters of GPUs.
• Experience working in an academic research environment and effective communication within such an environment.
• Ability to engineer industry solutions and maintain target service level agreements.
• Ability to work independently and quickly ramp up on complex, unfamiliar code.
• Experience with engineering practices, CI/CD pipelines, and Git.
• Experience contributing to and/or maintaining open-source projects.

Benefits

• Competitive salary range of USD $137,600 - $267,000 per year, with higher ranges for specific locations.
• Potential eligibility for additional benefits and compensation.