GS Caltex links SageMaker Canvas models to production pipelines
Start here
GS Caltex has connected models built by its process engineers in Amazon SageMaker Canvas to production workflows in Amazon SageMaker Pipelines, automating a handoff that previously required manual integration work. The South Korean energy and chemical company says ML engineers can now deploy models in a few hours instead of about a week. GS Caltex also runs batch inference every 30 minutes instead of keeping endpoints continuously running, which it says has meaningfully lowered inference costs. AWS published a technical walkthrough of the setup, with a reference implementation in a GitHub repository that uses a sample loan classification model and must be adjusted to each model's input and output formats. AWS advises using the approach only when the placeholder target does not affect feature values, and adding checks before production that file order and row counts match, stopping result generation if they do not.
Generated from reports · updated 35 min ago
Report timeline
Follow the reports to see the event from different sides.
- AWS Machine Learning BlogGS Caltex Production MLOps with Amazon SageMaker Canvas and Pipelines, Part 1
GS Caltex connects models built by process engineers in Amazon SageMaker Canvas to production workflows in Amazon SageMaker Pipelines, automating a handoff that previously required manual integration work. The South Korean energy and chemical company reports that ML engineers now deploy models within a few hours instead of about a week, and runs batch inference every 30 minutes rather than keeping endpoints continuously running, which it says has meaningfully lowered inference costs.
Heat for this event
Not enough continuous observations yet to draw a trend.