Business Problem
Huge amounts of product imagery data siloed in legacy on-prem applications. Each application contains about 250 million images and spans across geographies and multiple providers. Retrieval and analysis are slow and cumbersome.
Requirement
The solution needed to:
- Ingest at least 10,000 images (and their related analysis data) per hour.
- Be scalable to add new applications without changes to architecture.
Our Solution
VaporVM developed lightweight utilities, powered by Azure, capable of being serverless to connect to each application and ingest all images and documents. VaporVM provided the following services:
- Architecture design and development
- Serverless Functions for connectivity and orchestrations
- Software development for optimized ingestion (17,000 images per hour)
How we created value
- A centralized log repository for all imagery data, with single sign on. The system also provides image tracking to provide information of source application, use case and ingestion / update timestamps.
- The customer is now able to use a fully automated system to securely access up-to-date data without the need for external access and intervention
