Overcoming the Challenges of Experimentation on a Service Oriented Architecture

Growing from an early stage startup to a national leader in financial literacy is no small feat, and there are a ton of lessons that we have learned at Greenlight as we have grown. Long gone are the days where we would ship something and cross our fingers hoping that it makes some kind of impact on our customers. Now we’re in a world where we can learn ahead of time how much impact a feature will have on the business, before we even launch! In today’s conversation, we’ll discuss how we use Optimizely’s feature flags in our microservice architecture using Optimizely Agent while keeping user IDs and context synchronized.

This session will cover:

  • How we set up Optimizely Agent and use it in a kubernetes deployment
  • How we created a user-aliasing service
  • How we access Optimizely both on the frontend and in the backend services.
  • How to build a full stack feature
  • How to manage the rollout using Optimizely’s feature flags


Lupin is Greenlight's Director of Product Growth. She's a data-driven product manager that relies on experimentation to uncover new opportunities to improve conversion. Since she arrived at Greenlight a year ago, their registration conversion rate has increased by 40%.

Lupin Campos

Director of Product GrowthGreenlight Financial Technology


Stephen Schwahn serves as the tech lead for the growth team at Greenlight, where he pioneers the platform and methodologies that enable its growing experimentation program. Prior to Greenlight, he founded a music technology startup called Crescendo, where he embraced the build, measure, learn cycle.

Stephen Schwahn

Tech LeadGreenlight Financial Technology


Watch Now

Welcome back. Not you?


Sorry, there was an error. Please try again.