December 11, 2023

For many, ChatGPT and the generative AI hype train signal the arrival of artificial intelligence enter the mainstream. But despite little sea change in public awareness over the past six months, growing demand for AI could outstrip the infrastructure needed to power the myriad use cases that are emerging — and that’s exactly what Germany is doing. Startups quadrant Looking for address.

Founded in Berlin in 2021, Qdrant aims to provide AI software developers with an open-source vector search engine and database of unstructured data, an integral part of AI application development, especially as it involves the use of real-time data. Classify or tag.

To help its technology move deeper into the commercial space, Qdrant today announced $7.5 million in seed funding from lead investor Unusual Ventures, with participation from 42cap, IBB Ventures and angel investors including Cloudera co-founder Amr Awadallah .This is except 2 million euros Qdrant’s pre-seed funding ($2.2 million) raised last year.


For the uninitiated, vector databases are designed to store such unstructured data such as images, videos, and text, allowing people (and systems) to search unlabeled content, which is especially important for scaling use cases of large language models (LLMs) ) such as GPT-4 (which powers ChatGPT).

according to According to Gartner, unstructured data accounts for 90 percent of new data generated by enterprises and is growing three times faster than structured data. at the same time, the vast majority of AI research and development (R&D) projects have never made it to production, which Qdrant CEO and co-founder Andre Zayarni attributes to a lack of the right tools—finally, being able to connect LLM to real-time, unstructured data could be the foundation for anyone wishing to build There are plenty of opportunities for people with more useful AI applications.

“The vector database is a natural extension of their (LLM) capabilities,” Zayarni explained to TechCrunch. “GPT’s biggest limitation is that it only ‘knows’ events that happened before the model was trained, but if it is connected to a vector database, LLM’s virtual ‘memory’ can go through real-time and real-world data.”

Investors have also been paying attention. Just last year, a similar proposal to Qdrant called Pinecone garnered $28 million, though Zayarni sees Qdrant’s open-source foundation as a major selling point for potential customers.

“Engineers trust open source, and if there are OSS products that offer something similar or even better, it will be very difficult for proprietary software to compete in this market,” Zayarni said.

Of course, there are other open source players like Zilliz, a startup commercializing the Milvus open source vector database that raised $60 million last year. Earlier this month Chroma Received $18 million in seed funding Develop its “AI-native” open source vector database.

Qdrant has now raised $7.5 million in seed funding, which says something about what investors are thinking right now—any technology that promises to help advance AI and machine learning and extend its capabilities to all developers is clearly is an attractive proposition.

Qdrant spent the better part of a month fine-tuning its pitch deck for its seed round, Zayarni said, and received its first term sheet the day after the deck was sent out, and another term sheet two days later. .

“We’ve had over 20 VCs interested — almost all of whom want to join as co-investors later — and we’re likely to get more invitations,” Zayarni said. Their extensive experience and their model of being an active operating partner rather than just an investor was so attractive to us that we decided to partner with them.”

Today’s funding announcement comes months after Qdrant roll out it is Managed Cloud Offeringsdesigned to help developers with one-click deployments, automatic version upgrades, backups, and an upcoming database management interface.

With the fresh cash infusion, Qdrant is also developing an enterprise product that can be hosted on-premises or in a private cloud, which he expects to launch later this year, Zayarni said.