# Oh no, not another one! Databricks buys Electric > Source: > Published: 2026-08-12 11:51:33+00:00 ai/ml # Oh no, not another one! Databricks buys Electric Databricks has bought PGLite and Electric Sync Engine developer [Electric](https://www.databricks.com/blog/electric-joins-databricks-bring-wasm-postgres-ai-agent-sandboxes) to bring distributed state and real-time data synchronization between teams of agents, extending Databricks’ Postgres capabilities from the lakehouse to the edge. PGlite gives every agent its own lightweight Postgres where it runs, providing ultra-low latency access to local context, and Electric’s real-time sync engine synchronizes distributed state back to a central Lakebase, enabling teams of agents to collaborate without losing track of shared context. Electric was started up in the September 2021 by CEO James Arthur, who is listed as co-founder [along](https://electric.ax/about/team) with CTO Valter Balegas, CPO Kyle Matthews, and Chief Data Office Sam Willis. Kyle Matthews’ LinkedIn [entry](https://www.linkedin.com/in/kylemathews/) says he became ElectricSQL’s Chef Product Officer in May 2024, joining from being a full-time CTO at GatsbyJS. Balegas’ [entry](https://www.linkedin.com/in/balegas/ ) says he joined in March 2022. Sam Willis’ [entry](https://www.linkedin.com/in/samwillis/) says he joined in August 2023. An October 2023 Electric [blog,](https://electric.ax/blog/2023/10/10/welcome-sam) written by Purva Gujar who was hired by Electric in January 2022, says Willis joined as a founding engineer. A Willis quote says: “I came across Electric, and upon seeing a job opportunity related to it, I decided to pursue it.” The company has raised $7 million in funding from 18 angel investors and 5 VCs: Spark Capital, Lunar Ventures, StepFunction, Firestreak and Ninja Capital. Electric developed PGLite, a Web Assembly (WASM) version of Postgres that runs inside a browser tab or other application, and supports dynamic extension loading. PGlite began as a proof of concept by Neon co-founder Stas Kelvich. Electric developed that base and turned it into a database that millions developers use. It has grown from 1 million to 13 million weekly downloads in the last 12 months. The company has also developed three open-source products: Electric Agents - runtime for long-lived agents.Agents live as durable, synced entities — resumable across devices, observable across teams, forkable for review and experimentation. Electric Streams - data primitive for the agent loop. Persistent, addressable, real-time streams; a flexible swiss-army-knife for agent session data. Electric Sync - core sync engine technology. [Postgres Sync](https://electric.ax/sync/postgres-sync)is the read-path sync engine for Postgres, handling partial replication, data delivery and fan-out from Postgres to clients over HTTP. Databricks bought Neon in May last year in a $1bn deal, saying it would allow it to deliver serverless Postgres. It wanted to to strengthen its ability to serve agentic and AI-native app developers, as over 80 percent of the databases provisioned on Neon were created automatically by AI agents rather than by humans. Ali Ghodsi, Co-Founder and CEO at Databricks, said: “By bringing Neon into Databricks, we’re giving developers a serverless Postgres that can keep up with agentic speed, pay-as-you-go economics and the openness of the Postgres community.” It then [added](https://www.blocksandfiles.com/container-storage/2025/06/16/databricks-rolls-out-lakebase-postgres-and-agent-bricks-for-ai-era-apps/1592030) a Lakebase Postgres database layer to its lakehouse in June last year, enabling AI apps and agents to run analytics on operational data within the Databricks environment. The Electric Sync Engine, Databrick’s Neon unit [says](https://neon.com/blog/electric-joins-neon?utm_source=the+new+stack&utm_medium=referral&utm_content=inline-mention&utm_campaign=tns+platform), “keeps data continuously synchronized between a central Postgres database and connected clients like browser tabs, mobile apps, or agents running in a sandbox. Sync engines power collaborative products like Google Docs, Figma, and Notion, and the same technology is critical to keeping state synchronized across fleets of agents.” It adds: “A database that runs inside agent sandboxes and browser tabs, synchronized in real time with Lakebase Postgres in the cloud, lowers the entry cost for cloud apps and agents. Software abundance needs innovations like PGlite: millions of small, fast applications, each with real Postgres underneath.” Arthur and Balegas [blog](https://electric.ax/blog/2026/08/11/electric-joining-databricks) notes: “Lakebase is serverless Postgres for apps and agents. With scale-to-zero, native branching and support for both OLTP and OLAP. By combining Electric’s sync and local execution with the scalability and capabilities of Lakebase, apps and agents get a supercharged Postgres stack for optimal fast, reactive development, local query execution at machine speed and deployment at scale, for workloads of any scale. The Electric team will be joining Databrick’s Neon team to help build a complete backend for agentic apps. Neon says “The two teams are already at work on new backend capabilities, and we can't wait to see the apps and agents our customers build with them.” Every Electric open-source product will remain open-source. Terms of the acquisition were not revealed. **Bootnote 1** Databricks says AI agents don’t behave like traditional applications. Developers are moving from building traditional apps to building agentic applications, and the assumptions underlying traditional infrastructure are shifting with them. A traditional application has a known shape: its queries are written in advance, its data access patterns are predictable, and it runs in a place its architects chose. One managed Postgres database serves it well. The agents in agentic applications add a second surface. Alongside the durable, governed state every application needs, an agent generates a fast-moving context set. Agents differ in three ways that matter for a database: They decide what data they need at runtime, deciding their next move and updating their context several times a second. That inner loop wants data in the same process; the results it produces belong in a durable store. They run wherever the work is, often in sandboxed environments, where databases are only reachable via a network connection to the cloud. They work in groups, requiring fast local context and a shared, current view of what others have done to avoid duplicating work, acting on stale state, or arriving at conflicting conclusions. The Electric team built data primitives for the era of agentic applications. Lakebase delivers Postgres at production scale. PGlite brings WASM Postgres into the agent's own sandbox. Sync keeps the two consistent. By combining the power of Lakebase Postgres with the local execution of PGlite, agents get the lightweight databases they need to move fast and the infrastructure required to deploy at any scale. Bootnote 2 In the last 12 months Databricks has acquired Tecton, Mooncake Labs, Quotient AI, SiftD.ai, Antimatter, Panther, and now Electric.