AI-ready Vector Database Provider Qdrant Gets Backing

Engineers at the sea – Qdrant core team – © Qdrant

A database is not simply a database. The differences in kind have to do with how the data held is stored. The classic relational database is great when you have structured data. The newer graph databases don’t focus on the datapoint as the constituent unit but on the connections between them, making these useful for social networks or recommendation systems. The even newer vector databases are good for finding similarities within a mass of unstructured data, making them good for content-based retrieval, such as image or document similarity matching. For machine learning or AI, a vector database is probably the best bet. Qdrant is a Berlin-based company that has created an open source vector database. Open source doesn’t mean CEO and co-founder Andre Zayarni has no business model (they provide managed services and an on-premise edition), and so Qdrant has just raised $28M in a Series A investment round, as techcrunch reports – commenting that “unstructured data makes up around 90% of all new enterprise data”. High-profile early customers include Deloitte, Accenture, and Elon Musk’s AI venture Grok.