Story
Google Cloud
On HoldingOn Holding Uses Vertex AI for Model Serving
On Holding, a consumer discretionary organization in Switzerland, uses Vertex AI from Google Cloud to support applied AI delivery for ML and product teams.
Value results
| Category | Value result |
|---|---|
| Productivity | Handoffs in applied AI delivery sit in a shared queue instead of a mailbox trail |
| Risk and compliance | Vertex AI is the governed place ML and product teams use for applied AI delivery |
| Capability | New joiners can see how applied AI delivery actually runs |
Story
Inside On Holding, applied AI delivery used to depend on whoever still had the latest file. That pattern is common in consumer discretionary groups working out of Switzerland. ML and product teams needed a system that would still make sense after the original project team moved on.
On Holding uses Vertex AI from Google Cloud as the working layer for model serving. Google Cloud provides infrastructure, analytics, and AI services, including BigQuery and Vertex AI, for data-heavy and machine learning workloads. The practical change is simple: applied AI delivery has a home, and reviews happen there instead of in a forwarded thread.
Nothing in this writeup invents a savings number. What On Holding gets from Google Cloud is a durable place to run applied AI delivery and a way for ML and product teams to see the same model serving at the same time.
