Tesla’s Autopilot Machine Learning team is currently seeking a software engineer to design and build a database and associated RESTful API for command and control of a custom ML training accelerator (DOJO). In particular, this engineer will work with hardware designers, embedded software engineers, and ML engineers to design, prototype, and implement a web service that is responsible for interfacing to existing and future custom distributed systems for ML training. The candidate must be comfortable with from-scratch API design and associated back-end software stack specification. Expertise in design of web APIs and associated back-end database for eventual integration to a custom web application are essential for this position.
- Set technical direction, architect, implement and maintain backend API stack for Machine Learning coordination of a highly distributed system.
- Work on the platform of tools and infrastructure that the Machine Learning team needs to be effective. This spans the scope from machine firmware interface to frontend user interface.
- Automation and coordinate required hardware resources with the team managing the cluster hardware to maintain high availability.
- Interface with both Machine Learning and hardware design teams to understand current and future requirements and priorities. Requirements:
- Strong experience building APIs in Python, GoLang, Node.js, or Java.
- Experience working in multithreaded or highly distributed Linux environments.
- BS/MS in Computer Science or the equivalent in experience with evidence of exceptional ability.
- Minimum 5 years’ experience building web services.
- Interest in Machine Learning, Computer Vision or Neural Networks.
Extra but not required:
- Experience building modern web applications using React or similar component-based libraries.
- Experience working with large computer clusters.
- Experience working with distributed networks and IoT coordination.
- Experience working with workload management software such as Slurm or LSF.