Tyler Technologies is the largest SaaS vendor in the United States solely devoted to the public sector. The company’s Enterprise Permitting & Licensing solution is used by government agencies to automate and streamline community development and business management operations. Overall, Tyler’s products empower the public sector to create smarter, safer, and stronger communities.

Challenges: Skyrocketing costs, ad hoc deployment strategy, and limited opportunities for automation

When you visit the website for your local government to get a license, make a payment, or respond to a summons, chances are you’re using software developed by Tyler Technologies.

The leading U.S. software provider to the public sector, Tyler Technologies manages over 40,000 client deployments in nearly 13,000 locations across the country. Its cloud footprint is enormous, and as a significant one of the highest cost input to the business, it’s a footprint that Tyler is constantly trying to reduce. 

Each Tyler Technologies Enterprise Permitting & Licensing customer has a set of non-production environments, on non-production infrastructure, organized into “pods” and used to test Tyler software before deploying into production environments. This is a critical feature for Tyler, given the regulatory nature of its permitting and licensing services for government agencies: New features and bug fixes must be fully tested and vetted before pushing to production.

AWS, Tyler’s primary cloud provider, offers tools such as CloudWatch to help monitor and control costs, but the tools lacked the necessary automation to manage Tyler’s unique use case effectively. 

“We originally used CloudWatch alarms to power down non-production infrastructure outside of working hours,” explains Chris Camire, Senior Manager of Technical Services at Tyler. “But this wasn’t a long-term solution. Our clients often work late hours and need their non-production environments available on demand.”

The flip side of this challenge was that some cloud environments weren’t used daily, or even weekly, but could be needed on a moment’s notice. 

“As our client base grew, we continued to build additional pods to meet demand so that clients had secure, reliable environments to test in,” Camire says. “What we didn’t do was create a strategy for how we organized our clients across that cloud infrastructure. We had clients from different time zones, of all sizes, and with varying use patterns all sharing the same cloud deployments.” 

Solution: Cloud AutoStopping to improve visibility into and management of idle resources

The lack of manageability with existing tools led Tyler to explore more full-featured cloud cost management solutions, including Harness Cloud AutoStopping™, part of Harness Cloud Cost Management (CCM), for active management of idle cloud costs. 

“Cloud AutoStopping opened up new possibilities for cloud cost management,” Camire says. “We saw how reorganizing our deployments by geography, function, and use patterns could unlock game-changing savings.”

Tyler created a framework for reorganizing the cloud for cost efficiency – in their case, reorganizing the entire cloud estate by client time zone and client activity. 

“We started with time zones,” Camire explains. “Because our client base is public agencies with regionalized offices, we grouped clients that begin and end their workdays at roughly the same time.” 

When Cloud AutoStopping detects an idle state for all the clients in a group, it can automatically power down that instance until it detects activity again. “The cloud instance can re-start quickly, so that our clients don’t even know the difference,” Camire says. 

Results: Annualized savings of $1.2M after just six months

Organizing cloud instances around time zones was just the start. Using the Harness CloudAutostopping console, Camire and his colleagues gained new visibility into client usage – specifically, just how infrequently some clients were using their pod environments. 

“We identified clients that were idle for much longer periods, measured in weeks or even months, grouped those clients together, and set Cloud AutoStopping to shut them down accordingly,” Camire says. 

Grouping clients together by time-based criteria has produced amazing results. “Right away we were saving $15,000 to $20,000 a month, and in only six months, we saw our first savings topping $100,000 a month,” Camire says. 

Following the initial successful results, Camire and his team are working on establishing 100% coverage for Cloud AutoStopping rules, planning to replicate the cost savings throughout the cloud estate. Says Camire, “We’ll start exploring other Harness CCM features more in depth, like Cloud Asset Governance and implementing recommendations.”