How do you build a modern CI/CD pipeline for Salesforce?
Xometry's on-demand manufacturing platform makes it possible for customers to connect with suppliers and receive instant feedback on pricing, expected lead times and the manufacturability of parts. At the push of a button, the network delivers high-quality items within consistently fast lead times and at highly competitive prices. The company began working with Salesforce a couple of years ago after fast growth led to the need for more sophisticated internal processes.
As a two-sided marketplace, Xometry needs their Sales, Marketing, and Operations teams to have a 360-view of both their customers and suppliers. Instead of relying on multiple SaaS tools, the company was looking to benefit from the synergies of a single, cohesive platform. Salesforce was the preferred choice because of how easily it could be customized to suit their needs. But the amount of data Xometry generates from their various other systems has meant the integration work requires a huge technical effort.
We caught up with Chris Cope, Senior Director of Software Engineering, and Hongda Zhou, an all-round Awesome Admin, to learn about their experiences with Salesforce to date.
Painful manual processes
The team has been through a lot in the past two years. Compared to development in other areas of their platform, customizing their Salesforce instances was frustratingly slow and dependent on manual processes, as Chris explains:
For Chris, it was all just too painful:
Lack of visibility
Chris and the team had looked into building their own CI/CD toolchains based around both GitLab and GitHub, but the lack of visibility was a huge problem:
Stepping on each other's toes
The team of developers also soon found they were doing a lot of administration work catering to a growing number of users, and so Chris hired the team's first admin and trained up five of his software engineers in Salesforce development. But it wasn't until a few months back when Hongda joined and introduced the team to Gearset that things really improved:
A transparent pipeline thanks to Gearset
Chris was impressed with Gearset from the start. From his point of view, one of Gearset's major benefits is the visibility it gives the team into the current status of work being deployed along the pipeline.
As a tech company, Xometry's rate of innovation is an all-important metric. Whereas before, releases on Salesforce (with release announcements and downtimes) stood out 'like a sore thumb' compared to the rest of the company's platform, Gearset has changed all that, helping the company to view Salesforce as an effective development platform.
But it's not just the team that benefits from greater visibility, as Hongda points out:
The integration with Jira, in particular, is also a useful way of tracking progress and measuring performance:
Environments kept in sync
For Hongda, there are two other major benefits of Gearset: the ability to back promote metadata and the ability to seed sandboxes with real datasets from production for testing. For example, if Hongda has deployed new work downstream, developers working in upstream sandboxes can keep their environments in sync. And from the team's point of view, the ability to schedule deployments in advance is a big help:
Bringing modern DevOps to Salesforce
Asked whether he'd recommend Gearset to others, Chris makes an important point. Gearset helps bridge the gap between software companies familiar with the language of DevOps and contractors in the Salesforce ecosystem who are generally less experienced with DevOps tooling and processes:
Finally, Hongda is emphatic in his endorsement of Gearset, which he describes as offering a user-friendly way of doing DevOps for Salesforce: