About this role
Foundational ML engineering role to design, build, and optimize scalable training pipelines for multimodal model training and evaluation, working closely with data engineering and research teams. Responsible for implementing training strategies, automation and tooling for experimentation, and addressing performance bottlenecks to improve throughput and model performance.
Skills for this role
PythonPyTorchTensorFlowJAXDistributed trainingLarge-scale distributed trainingInference systemsScalable machine learning pipelinesMultimodal data processingText-image pairingVideo understandingSpeech-audio modeling3D data processingDiffusion modelsDataset optimizationData collectionData preprocessingAnnotationFeature pipelinesData loadersModel servingModel deploymentEvaluation frameworksExperimentation frameworksExperiment trackingHyperparameter tuningKubernetesRayAirflowMLflow""AWS""GCP""Azure