Many years ago, I regularly contributed to development for a local start-up that had won a contest for sizeable state grants. Some of the initial challenge—which was to convert a Meteor app to fully use graphql as its data layer—I have documented on the meteor page.

Beyond that conversion to a very large graphql schema unifying data stored in MongoDB and Postgres, I took on the priority of establishing a more distributed microservice architecture outside of Meteor to supplement it for more complex or long-running data tasks. Here, I set up RabbitMQ to power work queues and then established subscriptions up through the Meteor graphql layer onto the client.

In my role as a primary developer, I also frequently led sessions teaching other developers best practices, particularly around code structure and testing.