Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More
Timescale is looking to further advance its namesake open-source database platform with new AI capabilities announced today.
Timescale was founded in 2017 as a time series database (TSDB) technology based on the open-source PostgreSQL relational database. The combination of time series data and vectors has real value for enterprises, as it helps to enable generative AI applications with Retrieval Augmented Generation (RAG). That’s why Timescale this year in particular has been advancing its vector capabilities. In June, the company announced its pgvectorscale and pgai efforts, integrating advanced vector database capabilities with Timescale’s database platform. Now Timescale is going a step further with its new pgai Vectorizer developer tool that creates and syncs embeddings right in the database. As an open-source technology, pgai vectorizer can potentially be used by any PostgreSQL database user to help enable generative AI applications.
“We’ve taken this small idea of PostgreSQL for time series, and we’ve kind of grown into a much larger idea, built on our success there, which is PostgreSQL is the developer platform for any application,” Ajay Kulkarni, CEO and co-founder of Timescale told VentureBeat.
The intersection of time series data and vector database technology
The intersection of time series data and vector database technology is an area of focus for Timescale.
Kulkarni explained that these two data types are overlapping and can be used together in various applications. He noted that Timescale today has customers that use the database just for time series and some that use it just for vectors. A third category is customers that are starting to use the technology for both use cases. The intersection of time series and vector data allows for use cases that leverage both the temporal aspect of time series and the semantic capabilities of vector search.
Among Timescale’s early vector customers is electric vehicle startup Lucid Motors. Kulkarni explained that Lucid uses vector search on images that also have a timestamp, where the value of the images decays over time.
Kulkarni said that he sees the blending of time series and vector data as an important trend, where organizations are looking to leverage the strengths of both data types within a single database platform like PostgreSQL.
The goal is to simplify vector database management for AI
The new pgai Vectorizer is an extension of Timescale’s pgai effort that launched in June. The initial piece of that effort enables Timescale users to bring AI model integration directly into PostgreSQL.
The new pgai Vectorizer aims to streamline embedding management by making it as straightforward as traditional database operations. The open-source tool enables developers to create and manage embeddings across multiple text columns with simple SQL commands, automatically maintaining synchronization as underlying data changes. It also facilitates easy testing and deployment of different AI models, including switching between services.
The pgai Vectorizer builds upon Timescale’s existing vector database technologies, launched in June 2024. The company’s pgvectorscale extension is based on the open-source pgvector vector database extension. Multiple vendors including AWS and Google use pgvector to provide vector database capabilities to PostgreSQL
Timescale sees pgvector as having limitations at a larger scale, which pgvectorscale aims to address. According to Kulkarni, pgvectorscale provides improved performance and scalability compared to pgvector, while remaining fully compatible and open-source. He also argued that the open-source pgvectorscale can outperform other vector database technologies, including Pinecone.
Looking beyond RAG to agentic AI for vector database operations
Kulkarni emphasized that the pgai Vectorizer, just like the pgvectorscale extension, is open source and will remain that way. He hopes that by keeping the technology open source it will help grow the community of users and contributions as well.
Looking forward, the company sees pgai Vectorizer as part of a broader AI strategy.
“We’re essentially building RAG as a service right inside your database,” he said. “But we’re not stopping with RAG, we’re looking at agents.”