Timescale, a New York startup founded by Ajay Kulkarni and Mike Freedman, is on a mission to build open source database software for time-series data — i.e., data from internet of things (IoT) devices, financial platforms, and DevOps tools that is largely immutable and partitioned across both space and time. For the better part of four years, the company has made available TimescaleDB on GitHub under the Apache 2 license, alongside a managed version in private beta — Timescale Cloud — that enables developers to quickly deploy it to cloud environments, like Amazon Web Services, Microsoft’s Azure, Google Cloud Platform, and DigitalOcean.
Now it’s shifting its attention to the enterprise market.
Timescale today announced that it has raised $15 million in a series A1 financing round led by Icon Ventures, with participation from existing investors Benchmark, NEA, and Two Sigma Ventures. This brings the company’s total raised to $31 million, following a $12.4 million series A round in January 2018, and comes as Timescale introduces the newest version of its platform, TimescaleDB 1.2, with enterprise features like production deployment assistance and enterprise-level service level agreements.
Kulkarni said the capital will allow Timescale to expand its team as it accelerates growth.
“By combining the ubiquity of SQL, the reliability of PostgreSQL, and the performance needed for time-series workloads, TimescaleDB is uniquely positioned to become a critical part of organizations’ infrastructure worldwide,” he said.
TimescaleDB is powered by open source database management system PostgreSQL and implemented as an extension rather than a fork, which confers the benefit of out-of-the-box compatibility with PostgreSQL connectors like Periscope, Tableau, Zabbix, and Patroni. As an added advantage, it’s able to use PostgreSQL’s ecosystem for things like tooling, streaming replication, point-in-time recovery, and backups.
TimescaleDB creates continuous database tables — “hypertables” — across time-series data, which behave like standard PostgreSQL tables for queries, inserts, upserts, triggers, schema changes, and management. The abstractions are partitioned, or “chunked,” automatically across multiple dimensions and sized to optimize performance, enabling TimescaleDB to scale to tens of thousands of chunks and write millions of data points per second on a single node.
Timescale’s tools enable developers to join time-series with relational metadata within the database and to store schemas (or go schema-less) with JSON. On the operator side of the equation, they offer support for triggers and data archiving, in addition to access control, encryption, and automatic time-series data partitioning.
“Driven by the needs of modern web and IoT-scale time data, time-series databases are in the process of blossoming from a niche technology to a mainstream product category, equal in importance to transactional and analytic databases,” said Michael Mullany, a general partner at Icon Ventures. “[B]efore TimescaleDB, time-series databases were making unfortunate trade-offs that damaged their utility as a general purpose application platform. The enthusiasm we’ve seen for Timescale among its open source community, as well as its rapid adoption rate, make us extremely excited to back both the technology and this incredibly talented team.”
News of the funding comes a week after Microsoft’s acquisition of Citus Data, which develops an extension to PostgreSQL that effectively transforms it into a distributed database. Timescale also competes with startups like InfluxData, which raised $35 million in February.