Case Study: Deutsche Telekom scales enterprise AI agents with Qdrant

How Deutsche Telekom Built a Multi-Agent Enterprise Platform Leveraging Qdrant

Deutsche Telekom faced the challenge of scaling AI-powered assistants across its vast European enterprise. They needed to deploy chatbots and voice bots for customer service across 10 different countries, which required handling strict data segregation, real-time processing, and complex agent collaboration. They sought a platform-first approach and enlisted Qdrant as their vector database to help meet these demanding requirements.

The solution involved building a custom multi-agent Platform as a Service (PaaS) called LMOS, which was designed for high scalability and modular deployment. By integrating Qdrant, Deutsche Telekom gained a vector database praised for its operational simplicity, memory efficiency, and multi-tenancy support. This new infrastructure powered over 2 million conversations and slashed the time to develop a new AI agent from 15 days down to just 2.

Deutsche Telekom

Arun Joseph

Leads Engineering and Architecture