The following is the list of recent and relevant articles. Pinecone It was probably ahead of its time when, in 2021, it launched a vector database aimed at data scientists. The firm began to promote AI driven semantic searching as the use-cases started to take shape last year. As LLMs gain in popularity, companies are beginning to recognize the value of vector database.
Investors seem to be in agreement. Today, the firm announced that it invested $100 million into a Series B Investment based on a pre-valuation of $750,000,000. Edo Liberty said that in a conservative investment environment, these numbers were difficult to find. The company was growing rapidly and investors saw a chance to grab a market leader.
“We are the leaders and creators of this category. When we first introduced the vector databases category, nobody knew what we meant. This is a mature industry with many players. We are clearly ahead of incumbents, clouds and other players. Liberty told TechCrunch that it’s easy to bet on the leader of a category already formed.
Liberty says the first-to market advantage has helped them grow from a few clients last year to 1,500 customers today. The growth rate is similar to that of a consumer tool, rather than a database with a high level of technicality. The company has received interest from companies of every size, including technology companies like Shopify, Gong and Zapier.
“This is B2B tech that’s being adopted by consumers.” This is something I’ve never seen before. “You have to build these capacity faster, which is expensive and difficult,” said he.
He claims that the interest in vector databases has been sparked by LLMs, but this is not true. Both databases let you search large amounts data, but LLMs are more rigid since the data is baked-in. The vector database is built to support semantic searching but still has the flexibility of a database.
This whole knowledge-management system is much more flexible, efficient and simple to operate. [with a vector database]”, he declares. He uses GDPR for an example. It’s simple to delete records from a database. Because of the structure of a model, it’s much more difficult to remove incorrect information.
Peter Levine sees vector databases as an important part of the AI stack. He will be joining Pinecone’s Board and is leading today’s investment for Andreessen Horowitz. “We think that Pinecone and the vector database have the potential to be an important component of a new AI data stack. Levine told TechCrunch he was “very confident” that putting resources before the company would help them reach Edo’s Vision.
Levine believes that vector databases can be used with LLMs to provide truth, which would reduce the hallucination issues we’ve seen. “Well, they do work together. “I mean think of LLM almost as an application that sits in this database. The database will store information and feed that to LLM so it can provide more accurate results and long-term data storage, he said. Liberty views the database as long-term storage that LLM can use.
The company plans to hire with a runway worth $100 million. Liberty has 100 employees now, but expects to have 150 or 200 by the end of the year.
The vector database space has been heating up since Pinecone launched a few years ago with players Qdrant, Zilliz and Chroma, all raising funds recently.
Andreessen-Horowitz led the investment today, with ICONIQ Growth joining him as well as previous investors Menlo Ventures & Wing Venture Capital. The company has now raised $138 million including a $28 million Series A last year and a $10 million seed investment in 2021.