TIMESCALE

Ajay Kulkarni and Mike Freedman (a Princeton University CS professor) were building an IoT analytics platform and hit a wall: existing databases couldn’t handle the volume and velocity of time-series data efficiently. Time-series data — measurements that change over time — has unique access patterns that general-purpose databases handle poorly.

In 2015, they pivoted from the IoT application to building TimescaleDB — an open-source time-series database built as a PostgreSQL extension. This was genius: developers already knew SQL and Postgres, so adopting Timescale required zero new learning.

Open-source core (TimescaleDB extension) with a managed cloud service (Timescale Cloud). The cloud service charges based on compute, storage, and data compression.

Enterprise self-hosted licenses add high-availability and support. The PostgreSQL foundation means customers can use Timescale alongside their existing Postgres tools and workflows.

TimescaleDB — PostgreSQL extension for time-series data. Timescale Cloud — managed database service.

Continuous aggregates — automatic real-time rollups. Compression — up to 95% storage reduction.

Tiered storage — automatic data movement between performance tiers.

PostgreSQL ecosystem leverage. By building on Postgres, Timescale tapped into the largest open-source database community.

Developers could add time-series superpowers to their existing Postgres databases without migration. Strong developer content and documentation drove organic adoption.

InfluxDB is the other major time-series database. Amazon Timestream competes on AWS.

General-purpose databases like Postgres are getting better at time-series workloads, potentially reducing the need for a specialized solution.

Tiger Global led the Series C at a $1 billion+ valuation. Earlier investors include Benchmark, NEA, and Icon Ventures.

Total funding of $181 million.