Using the right tool for the job to get models from development to production with ease. 

1. Data Ingestion from object storage 

2. Dataset preparation (infer labels, splitting, augmenting, optimizing) 

3. Model Development from scratch and Training Strategies (one device, mirrored, multi-worker mirrored) 

4. Model Performance Hyperparameter Tuning strategies (RandomSearch, Hyperband, BayesianOptimization, Sklearn) 

5. Model Serialization to object storage 

6. Prediction Sampling 

7. Automate model serving: 1) containerize the model and inference engine and 2) containerize the inference engine and post requests to the model 

8. Interact with the model by upload samples via user interface 

9. Performance visualization (logging, metrics, dashboards) 

The demo covers several topics across the lifecycle for extending Kubernetes to perform common data science tasks from data ingestion to inference monitoring.

Local ODSC chapter in NYC, USA

Instructor's Bio

Cory Latschkowski

Senior Specialist Solutions Architect at Red Hat

Experienced platform architect with a demonstrated history of working with large, complex systems and translating new technologies into practical solutions. Strong Linux and open source background with practical DevOps experience. Digital and cultural transformation evangelist.

David Marcus

Principal Specialist Solution Architect at Red Hat

North American Lead Associate Principal Solutions Architect for Red Hat in the domains of Data Science, Artificial Intelligence, Machine Learning and Edge Computing. I hold a Bachelor's Degree from Kennesaw State Univ. and Masters Degree from Pennsylvania State Univ. in Computer Sciences. Prior to my career at Red Hat, I work for Lockheed Martin notably for Space Systems Co. contracted on 3 human space flight missions for NASA's Orion Program.


  • 1

    ON-DEMAND WEBINAR: “The right tools for the job” A way to do end-to-end Machine Learning

    • Ai+ Training

    • Webinar recording

    • Welcome to ODSC East 2023 in Boston or virtually!