Serve’s Machine Learning (ML) platform is an important core part of our autonomy. It empowers us to train and test all kinds of ML models for various real-world tasks. We also use it to mine useful data in terabytes of sensor recordings that we capture every day.
We are looking for an engineer who will join our Machine Learning Infrastructure (ML Infra) team on a mission to build and improve this platform. We use Apache Beam (Dataflow) for our pipelines, Bazel as our build system, BigQuery via dbt, MongoDB and GCS for storage, Kubernetes for service deployment, and Airflow for our workflow management.
Develop and maintain ML infrastructure, such as sensor data ETL pipelines, hard example data mining, continuous training pipelines, annotation platform, etc.
Develop MLOps system for managing lifecycle of ML cloud training and inference as a service pipelines. Continuously improve ML model development, management and deployment processes.
Work together with ML engineers, design metrics for ML tasks to mine sensor data of interest.
Design and implement algorithms, such as collaborative filtering, active learning, etc., to rank/score annotation candidates.
Work with annotation provider on setting up the annotation process, quality control and feedback loops.
Make sensor data and its derivatives widely discoverable and accessible for Robotics Engineers across the entire company.
BS or MS in computer science with focus in data engineering and large scale ML systems
3+ years of industry experience building, running and improving large-volume ML training and validation pipelines.
Experience with building native cloud applications
Experience with large scale data processing pipelines in production.
Proficient in at least one of the following languages: C++, Python, or Go.
Hands-on experience and good knowledge of Computer Vision and Deep Learning.
Strong tendency to automate own and others’ workflows.
Experience with data discovery and visualization tools like Voxel51, Facets
Experience with database systems like BigQuery, MongoDB
Experience with Nvidia Jetson platform, e.g. CUDA, TensorRT, etc.
Experience with Big Data products such as Apache Beam/Spark/Hadoop, GCP BigQuery, AWS Redshift.
Hiring locations
Currently, we are only hiring US and Canada residents, even for remote jobs.
More about us
Serve Robotics is the team that created the Postmates delivery robot and brought it to life in LA, completing tens of thousands of deliveries in LA’s busiest neighborhoods. Now we're an independent company growing rapidly in order to connect people with what they need via robots designed to serve people.
We are proud to be an equal employment opportunity and affirmative action employer. Qualified applicants are considered without regards to race, color, religion, national origin, age, sex, marital status, ancestry, physical or mental disability, veteran status, gender identity, or sexual orientation.
See Serve in action
Serve at TED 2020 main stage
Slack
Barracuda
Yahoo
