- Design, build, and maintain scalable pipelines that span data ingestion, annotation, validation, training, evaluation, and deployment, ensuring reproducibility, consistency, and traceability across the full ML lifecycle.
- Build and integrate annotation workflows with upstream data ingestion and training systems, enabling efficient task creation, labeling, QA, and dataset updates that directly support model iteration.
- Analyze model performance and failures, and drive targeted data improvements by connecting production signals, data mining, and annotation workflows into continuous feedback loops.
- Implement systems for experiment tracking, dataset versioning, and model lineage to enable reliable comparison and iteration across experiments.
- Develop and maintain CI/CD workflows tailored to ML systems, enabling automated testing, validation, and deployment of models and pipelines.
- Collaborate with embedded and platform teams to support the deployment of models to edge environments, ensuring compatibility, performance, and reliability.
- Implement monitoring, logging, and feedback systems to track model performance in production and drive continuous improvement through data and model iteration.
- Optimize training and inference workflows across cloud environments, including efficient utilization of GPU and compute resources.
- Work closely with applied scientists, embedded engineers, and data teams to ensure alignment across data workflows, model development, and deployment systems.
- Participate in and improve the full ML lifecycle, from raw data ingestion and annotation through training, evaluation, deployment support, and post-deployment analysis.
AWSDockerPython+10 more