Qdrant launches first vector database as a managed hybrid cloud

Today’s complex, unstructured data — text, images, audio and video — are difficult for traditional databases to handle. They struggle with this complex, demanding data with large numbers of variables and features. 

Vector databases have emerged to handle this problem; they are specialized databases that can efficiently index, query and retrieve the high-dimensional data critical to building and running AI. 

However, the technology is rather new, and deployment options are slim. Enterprises want to make the most of vector databases without sacrificing control of their data. 

Today, organizations have a new option with the launch of Qdrant Hybrid Cloud. Open-source vector database Qdrant says it is the industry’s first vector database to be offered in a managed hybrid cloud model. 

“Vector databases are designed to handle complex high-dimensional data, unlocking the foundation for pivotal AI applications,” Qdrant CEO Andre Zayarni told VentureBeat. The company’s new hybrid cloud service “enables businesses to explore new AI use cases that they would otherwise not be able to unlock with third-party cloud services.”

Vector databases across any cloud, on-premises or the edge

The vector database market size is expected to nearly triple from $1.5 billion in 2023 to $4.3 billion in 2028. Numerous providers have emerged in the space along with Qdrant, including Pinecone, MongoDB, Milvus, Rockset and others. 

Qdrant says it sets itself apart with its dedicated database tailored for high-dimensional vector data. Now, its Qdrant Hybrid Cloud allows customers to run vector search workloads in their own environments, ensuring data stays with them.

“We designed the offering to give our users and customers the maximum control and sovereignty over their data and vector search workloads across any cloud provider, on-premise, or edge location,” said Zayarni. 

Qdrant Hybrid Cloud powers several use cases, including generative AI and retrieval augmented generation (RAG); enterprise search (semantic, similarity or neural search); personalized recommendation systems; and data analysis and anomaly detection. The service is supported in any Kubernetes cloud environment and a growing number of other platforms. 

“AI excels in extracting meaningful insights from vast amounts of unstructured data, enabling enterprises to understand and utilize this information with advanced context and precision,” said Zayarni. 

Flexibility to run in any environment

One important use case is an enterprise developing an internal AI knowledge assistant/chatbot to support specific business functions such as sales or R&D, said Zayarni. This AI assistant requires access to internal documents to provide relevant results — but this can include sensitive proprietary data that must stay within a company’s security perimeter for compliance and data privacy reasons. 

The risks of not having full control of data are dire; these include technical risks such as infrastructure inflexibility. Organizations also can’t optimize costs because they don’t own the infrastructure on which they’re running, and they also can’t run database workloads close to the core application (among others), said Zayarni. 

“Enterprises need the flexibility to run their vector database applications in any environment with full control over their data,” he said. Qdrant Hybrid Cloud “takes the next step in enabling large enterprises to face complex challenges and better build and implement robust, next-gen AI applications while meeting strict risk and compliance standards.”

Flexible deployments help ensure privacy, security, compliance

Enterprises are shifting from prototyping AI tools to actively deploying them into production. In this new phase, Zayarni said, privacy, data sovereignty, deployment flexibility are top of mind. These are all critical to developing, launching and scaling new apps, whether customer-facing services such as AI assistants or internal tools for knowledge and information retrieval or process automation.

Vector databases are a relatively new technology, he pointed out, and deployment options “are at the beginning.” Up until now, enterprises have had few choices — typically either open-source deployments on their own infrastructure or the use of ready-made managed services. 

Vector databases haven’t yet been offered in a managed hybrid cloud model because developing this type of service is “a higher effort,” and core products must be compatible, said Zayarni. However, flexible deployments help ensure privacy, security, compliance and cost-efficiency. 

“This is where we see the market going now and why the hybrid cloud deployment option becomes very important,” he said. 

Ultimately, vector databases “represent a new frontier in data management, in which complexity is not a barrier but an opportunity for innovation,” he added.