Neon previews serverless PostgreSQL database service
Startup emerges to launch a new developer-focused cloud database as a service based on open source PostgreSQL, with capabilities to enable branching for application development.
Startup Neon Inc. introduced a new open source cloud serverless PostgreSQL database focused on developers.
The San Francisco-based vendor released the serverless service in a technical preview on June 15, after originally intending the system to be for general availability. Neon said it will make the service generally available in the coming months.
Neon is led by Nikita Shamgunov, who is known in the database community as the founder of unified database vendor SingleStore, formerly known as MemSQL. Shamgunov is also a partner at venture capital firm Khosla Ventures, which has also invested in Neon. To date, the startup has raised $24 million.
Getting PostgreSQL to work well in the cloud is a challenge that multiple vendors besides Neon are trying to take on.
Among the cloud-managed PostgreSQL services are Google’s AlloyDB, Amazon RDS for PostgreSQL, EDB's BigAnimal cloud database and Azure Database for PostgreSQL from Microsoft.
Neon is looking to differentiate itself with its serverless PostgreSQL database service by focusing on developers and integrating tooling to build applications.
A key feature of Neon is the ability to branch the database in order to try out new scenarios and build additional capabilities. Unlike some PostgreSQL services, Neon does not yet support analytical data and is intended for only transactional data.
Simulating supply chain changes with a cloud PostgreSQL database
Among the early users of the Neon database is Giancarlo Chiappe, senior software development engineer at Atlanta-based supply chain management system vendor KetteQ.
KetteQ is now developing the next-generation version of its product, which provides users with supply chain simulation capabilities for their data. In this architecture, KetteQ plans to use Neon for transactional data and the open source Clickhouse database for analytical data, Chiappe said.
With the ability to branch the Neon database, users will have the capability to change forecasting values on the branched Neon database and see the impact in real time, he said.
For example, a user in a simulation could shut down a factory and see the impact on the supply chain. The forecasting capability that Neon enables will lead to better supply chain decisions for businesses, Chiappe said.
Giancarlo Chiappe Senior software architect, KetteQ
"In this next-gen version of our product, Neon will be useful for its managed instant up database and branching features," he said. "Internally, we will start new Neon databases and branches on the fly to provide a secure simulation environment for the client where the original production data is not affected."
Developer spin to PostgreSQL cloud database deployment
After spending a decade building SingleStore, Shamgunov said he still sees opportunities in the database market.
Shamgunov said his goal with Neon isn't to build a competitor to SingleStore, which he noted is focused on data-intensive applications. Rather, his goal is to build a serverless PostgreSQL database that supports what he said he sees as an underserved market: developer-focused application development.
"We're building publicly as open source, and we're building for the developer with a focus on developer experience," Shamgunov said. "We support developer workflows, so we are helping people to build an app, rather than just providing infrastructure to run the app."
As a serverless PostgreSQL service, a developer can get a Neon database up and running in a few seconds, Shamgunov said.
Neon also supports continuous integration/continuous deployment (CI/CD) workflows in which developers constantly iterate on application code. A common feature of CI/CD workflows is the ability to maintain a branched environment that a developer can use for building and testing a new feature, without disrupting an existing branch of production code.
"We want to encourage people to branch a PostgreSQL database with every code deployment," Shamgunov said. "Developers can branch the database, run tests and verify schema to make sure that whatever changes have been made in the database are robust."