Senior Software Engineer, AI Infrastructure
Covariant
AREAS OF FOCUS
- Building services and APIs to search and annotate our rapidly growing robot dataset
- Designing libraries to help us train, deploy, monitor, and understand our models
- Full stack development of tools that leverage our libraries and services to visualize and explore Covariant’s robot data
YOU WILL
- Work closely with the research and solutions teams to spec, develop, and ship features for our robot data platform
- Lead and manage full-stack projects with cross-functional stakeholders
- Build tools to search and visualize robot telemetry data and facilitate fast performance iteration
- Implement scalable data pipelines to ingest and process robot telemetry data
- Develop and deploy distributed systems that span customer warehouses to the public and private cloud
- Advocate for and facilitate quality software design principles including system observability and debuggability
YOU HAVE
- 4+ years of programming experience in modern programming languages such as Python
- 4+ years of experience working on full stack, backend web development, or cloud infrastructure
- Designed, built, and deployed modern web APIs
- Designed and deployed solutions using public cloud providers like AWS
- Experience with containerization technologies like Docker and container orchestration platforms like Kubernetes and Amazon ECS
- Strong communication skills; able to efficiently communicate technical details to a varied audience
- Experience with building model training infrastructure, libraries, and tools
- The ability to work independently on open-ended cross-functional projects
NICE TO HAVES
- Experience architecting data infrastructure for machine learning systems
- Experience with Django and/or Postgres
SAMPLE WEEK IN THE LIFE
- Develop a scalable data pipeline leveraging services like Amazon SQS or Kinesis
- Design a new database model and corresponding API endpoints and views
- Deploy a service to Kubernetes and monitor its performance
- Triage and debug a performance issue in Postgres
- Add a feature to a computational graph library
- Meet with the research team to gather requirements and understand how we should support a new research project, such as training and deploying a new model
- Prepare a technical deep-dive presentation on a project you recently completed
- Independently run a meeting for your latest project to keep stakeholders on other teams up-to-date