PostgresML, the AI Postgres extension, announced the general availability of its end-to-end machine learning operations platform.
PostgresML allows developers to prototype and deploy AI applications on PostgreSQL by bringing the latest machine learning and large language models directly into the popular relational database – reducing complexity for app developers and yielding a host of additional performance, cost and quality advantages.
The problem with AI Infrastructure
Developing AI-powered applications requires a range of microservices to carry out the entire machine learning workflow, such as data preparation, inference, and fine tuning along with a stateful database to store the features.
This standard approach to MLops requires significant overhead, with some companies hiring experienced ML engineers to build and maintain the custom infrastructure, and teams of data engineers to maintain pipelines. Additionally, the required network hops from one service to another increase latency and require potentially unsafe data transfers.
“The recent explosion in AI power has only driven the costs and complexity for application developers higher,” said Montana Low, CEO and Co-Founder of PostgresML. “By streamlining the infrastructure requirements, we enable developers, founders and in-house teams to concentrate on creating intelligent and engaging applications.”
The Postgres Renaissance
“PostgresML is at the forefront of a trend wherein widely adopted developer tools and databases (such as Postgres) are being adapted and extended to an AI-first world,” said Sarah Catanzaro, General Partner at Amplify Partners in their funding announcement. The seed round was led by Amplify Partners with Angels participating in the round including the co-founders of Instacart and founding investors from Vercel and Box Group.
“Postgres is one of the most beloved developer tools on the planet. As such, it’s frequently the first database folks get started with to run their production applications. Bringing ML closer to where companies already store their data (i.e., Postgres) makes it easier to get started,” Catanzaro said.
PostgreSQL’s scalability, cost-efficiency and longevity as a nearly 40-year old RDBMS has also lent itself to the trust of enterprises such as Apple and Cisco over incumbents such as AWS, RDS or Aurora. According to The State of PostgreSQL 2022 report, usage is growing, with SMBs using Postgres more than they did one year ago, and the majority of respondents of all sizes (55%) saying they have increased their use of the database.
About PostgresML
The company, which was founded in 2022 by former Instacart ML infrastructure engineers, has gained significant developer traction, amassing several thousand GitHub stars since launch.
PostgresML users can deploy AI applications such as chatbots, forecasting, search, fraud detection and perform novel generative AI techniques such as retrieval augmented generation quickly and at scale. They can use simple SQL or Python and Javascript SDKs – making AI more accessible to developers without classical ML training.
“We’ve seen firsthand how the complexity of ML infrastructure can hinder the adoption of AI-driven applications,” said Catanzaro. “This is huge, especially for those smaller companies who view machine learning as a walled garden.”
With the funding and general availability at launch, PostgresML is continuing to evolve the platform to handle more sophisticated enterprise workloads.
PostgresML is an end-to-end MLops platform in a simple extension for PostgreSQL. It enables users to build fast, simple and powerful ML/AI models right inside their database using the simplicity of SQL.
View source version on businesswire.com: https://www.businesswire.com/news/home/20231109984513/en/