Case Study: Pit Rho achieves 70% server cost savings with Spot (Spotinst)

Using Spotinst As The Capacity Management Tool For Docker Environment

Pit Rho, a data- and machine‑learning services firm, needed an extra ~16 CPU cores and 32 GB RAM every race weekend but didn’t want to pay for peak capacity year‑round. Their prior approach—buying larger reserved machines—was costly and constrained real‑time processing. Pit Rho adopted Spot’s automated Spot Instance management to meet intermittent peak demand while keeping their infrastructure based on Docker and Rancher.

Spot integrated with Pit Rho’s Rancher-managed containers so new worker machines automatically scale up, join RabbitMQ/Postgres/Redis, and begin processing with no downtime; Spot handles bidding, provisioning, and replacing expired instances. The result: about 70% cost savings versus On‑Demand and nearly 50% savings versus reserved instances, with seamless scaling and no long‑term commitments thanks to Spot.

Pit Rho

Gilman Callsen

Founder & Chief Technology Officer