Replicated: Leading a Lean Team with Lean Infrastructure and Testing

Replicated is one of the standouts of the LA tech startup ecosystem, having rapidly and capital-efficiently built a technically complex product for SaaS engineering teams everywhere. Many tech startups face bloat and massive growing pains as they shift gears; Replicated’s commitment to smart growth shows that this common bloat is far from inevitable.

Frank Shotwell, Replicated’s Director of Engineering, leads an engineering effort that’s long in ambition: Replicated provides the tooling and infrastructure for SaaS solutions to deploy on-prem without needing to build the effort from scratch, giving SaaS providers access to the coveted enterprise market. This isn’t simple consumer tech or eCommerce: Replicated is complex, deep technology that requires great talent to build it.

Frank recruits engineers from the incredibly competitive pool of the booming Los Angeles tech ecosystem. “I always aimed to have a focused team of more talented people: building a great deep-tech product like ours isn’t something you can throw bodies at and hope to get the quality you need. Hiring in LA makes that even more true: we’re going to always gear ourselves to be lean and take pride in that.”

A Focus on Product

Frank’s leadership of his team goes beyond managing sprints. Too often, scaling engineering teams find themselves bogged down with commitments to maintaining infrastructure, devops, testing, and monitoring as they cross the chasm from scrappy startup to juggernaut: these commitments degrade average developer velocity and force teams to hire to compensate. Frank has taken seriously his technological leadership, integrating fresh-to-market technologies that minimize the infrastructure, and devops burdens on his team.

It wasn’t clear that the market had a technology Frank could use to repeat that success in testing.

“We make unit testing part of basic development practice here at Replicated, but I didn’t want to saddle the team with maintaining front-end tests against a rapidly-changing UI and knew that scaling an internal QA engineering department would eat up resources and add management overhead to everyone on the team.”

Frank saw that machine learning was driving a shift in the testing market, and chose ProdPerfect because he saw that E2E testing driven by data could not only save him resources, but also provide meaningful, targeted quality assurance without slowing his deployments.

“Within a few weeks we were able to move from ad-hoc regression testing to automated testing that we can run early and often. We catch multiple bugs every month that would have hit prod and could have been painful, and we use the test result videos to quickly diagnose any issues. We’re fixing bugs faster and thus actually increased our developer velocity. And I can use my resources to make my next great product hire.”

Frank’s loving the results and has made ProdPerfect part of his suite of scaling technologies.

“With ProdPerfect, I feel like we have a peripheral team whose attention and effectiveness is what I would expect from our own engineers. I know we are getting far more responsiveness and coverage than we would otherwise be paying for.”