INDY Cinema Group

The Problem

INDY OS is a comprehensive operations platform for cinema owners, from independent theaters to large chains. It handles booking, scheduling, showtimes, ticketing, website management, inventory, employee scheduling, and more. The platform was deployed across the UK, New Zealand, Australia, the US, and Mexico, serving hundreds of cinemas.

When I joined, the engineering team of eight was moving fast but without much structure. Deployments were manual, took over an hour, and had to go out to multiple regions around the world. There was no standardized CI/CD, limited observability, and communication between engineering, product, and support was unstructured.

What I Did

I worked across three fronts:

1. Developer experience and automation

I partnered with DevOps to automate the entire multi-region deployment process, turning a one-hour manual process into something reliable and fast. We introduced CI/CD with GitHub, running automated tests on every pull request. We created a staging environment where daily end-to-end test runs on critical features. We added Slack integrations that notified the team on every deploy with a direct link to the diff, and tools to monitor code coverage to drive test adoption.

2. Observability and reliability

We implemented New Relic and structured all logging into dashboards that gave us clear visibility into application health across regions. We set up alerting around deployments so issues could be caught fast. The platform was serving millions of requests per minute, so having confidence in every release was critical.

3. Team operations and cross-functional communication

I introduced structure to how engineering worked with product and support. That meant improving the on-call system, how issues were escalated and triaged, and how priorities were communicated. The goal was to create enough process to be reliable without slowing the team down.

The Result

The team could deploy daily with confidence to multiple regions. Engineers spent their time on features and important work instead of fighting deployments and debugging in the dark. The platform stayed reliable at scale across regions.