Cloudify’s SaaS product and services are built on complex, engineering-heavy development. From continuously improving their solutions to keeping up with the constantly changing public cloud architectures and replicating client issues in their own cloud environments, Cloudify depends on significant compute resources. This young, early stage innovator must guard their limited resources carefully, making sure not to waste money on over- or under-provisioning their public cloud compute and leveraging any available cost reduction opportunities.
Commitment-based discounts offer the richest savings available, but they require accurate forecasting over 1- and 3-year terms – practically eons for a small, growing start-up. Cloudify’s varied usage and need for flexibility across regions and machine types make it very difficult to do the forecasting necessary to take advantage of those commitments.
Cloudify was left to rely on manual management of compute discounts and all the forecasting that these required. This meant they didn’t always get maximum value from those efforts — usually because the commitments were tied to specific machine types that were deprecated or because the need for certain workloads was too unpredictable to commit to a set amount of money or usage. For a young, growing start-up company, this meant less money was available for reinvestment in the product, thus necessitating an automated solution that could optimize their compute costs without sacrificing their infrastructure flexibility.