Timescale has announced the launch of their Dynamic PostgreSQL product for business applications to more efficiently utilize their database compute and storage resources, even with variable or bursty workloads.

From the company that was first to supercharge PostgreSQL for time series, analytics, and AI, Dynamic PostgreSQL is Timescale’s unique twist on PostgreSQL managed services for relational workloads. It allows businesses to pay only for the compute and storage they use, rather than forcing them to allocate and pay for these resources at their peak, as is common with managed database services based on static resource allocation.

By allowing customers to choose a minimum and maximum CPU range, customers have the flexibility and scalability of serverless while solving the common problem of serverless offerings, which often surprise customers with exorbitant bills and underpowered, slow experiences.

The news comes after Timescale’s recent introduction of usage-based storage to their time-series cloud product. The feature removed the need for customers to calculate and predict their storage allocation and ensured they only pay for what they store—saving existing customers an average of 35% storage cost in the first month.

Dynamic PostgreSQL for relational workloads includes usage-based storage and introduces dynamic compute, which allows customers to choose a CPU range, but only pay for actual usage so they can “buy the base and rent the peak.” This flexibility ensures cost predictability without impacting performance.

Timescale’s Dynamic PostgreSQL product is available now.