As a ML Platform Architect on the Central Software team, you will play a vital part in scaling machine learning (ML) efforts across Boston Dynamics.  In this role, you’ll design the tools, infrastructure, and pipelines that unify how our many product ML/RL engineers and researchers get stuff done. You’ll work at the forefront of robotics innovation, using your skills to empower teams shipping new robot capabilities. 

How you will make an impact:

• Own the overall design and long-term growth of an ML platform, including data management, compute for model training, deployment, and optimization.

• Work closely with ML/RL contributors across teams  to understand and prioritize needs

• Develop and propose strategies, leveraging on-premises and cloud-hosted resources, to address compute, storage, and networking capabilities for fast/flexible ML capabilities

• Propose and communicate roadmaps, schedules, and design concepts.  Align stakeholders’ direction and priorities

• Work closely with ML Platform engineers to implement, deploy, and maintain designed infrastructure

• Serve as a technical resource to stakeholders across BD pursuing ML-enabled solutions

• Work closely with engineering leadership to identify and answer critical product questions

We are looking for:

• 8+ years experience as a ML platform engineer, Senior SWE / technical lead, or equivalent

• Experience effectively working in a large multi-language monorepos, including build technologies

• Proven track record developing, debugging, and deploying complex applications in Python, Julia, C++ or equivalent

• Demonstrated proficiency using ML libraries/frameworks (PyTorch, Tensorflow, Keras, Pandas, Numpy)

• Experience using hybrid infrastructure, combining on-prem and cloud (GCP, AWS, Azure) resources, for ML applications

• Demonstrated experience with scalable ML deployment methods (Docker, K8s, Ansible, Terraform)

• Experience with storage and data transformation technologies (S3, GCS, Airflow, Spark)

• Experience with Agile, Scrum or other lean methods.  Be able to work closely with high performing engineering teams

• Ability to work collaboratively in cross-functional teams

• Masters in Engineering, Computer Science, or other technical area

#LI-JM1