Case Study: Major Leading Automotive Company achieves fast cloud access and scalable high‑accuracy image labeling with Softeq

Machine Learning-Enabled, Supports Remote Access Through Cloud Storage

The client, a Germany-based developer of software products for major automotive companies, faced slow client-server communication and frequent delays because their ML-enabled image/video labeling ecosystem and media storages were hosted on on-premises servers in Berlin while labeling teams worked remotely. To resolve this, they engaged Softeq to migrate the labeling web application and media storage to the cloud and to redesign the system to better support manual annotation, ML-based preprocessing, and remote access.

Softeq delivered a standalone, role-based web app (Admin Panel and Workspace) with CNN-based image preprocessing, migration to Azure Cloud, and an architecture split across services such as Redis, RabbitMQ and Docker to improve scalability and performance. After migration and architecture refactor, the system reported faster, more reliable remote access and passed load testing for concurrent labeling teams; the customer has since contracted Softeq for phase two feature additions including HDR and real-time video annotation.