Leading the team and the release process

Context

At Event Network I was Senior Lead Magento 2 Developer for 18 Magento 2 storefronts on a shared codebase, connected to ERP and point-of-sale systems. I led a team of about 5 developers.

The challenge

With 18 stores on one codebase, a single change can reach every store at once, and every store depends on integrations that keep in-store and online inventory, orders and fulfillment in sync. The day a deploy broke checkout across the storefronts made that risk concrete.

What I did

• Led the team. A team of about 5 developers working on the shared Magento 2 codebase.
• Owned the release process. The Git workflow, pull-request review and approval, and every deployment to production.
• Built a checkout regression suite. Selenium driving Chromium through cart, checkout and payment against staging before every deploy.
• Added monitoring so failures alert the team, not customers.
• Built the ERP/POS backbone. The integrations that synced inventory, orders and fulfillment between the shops and the websites, on Magento's REST and GraphQL APIs, kept running through Magento upgrades and changes on the ERP and POS side.
• Kept the platform current. Custom modules and themes, multi-store configuration, upgrades and security patches, and performance with Varnish, Redis, Elasticsearch and a CDN.

Stack

Magento 2, PHP, MySQL, Git with pull-request review, Selenium with Chromium, REST and GraphQL APIs, ERP and POS integrations, Varnish, Redis, Elasticsearch and a CDN.

Outcome

From then on, checkout regressions were caught on staging before a deploy, not by customers, and production failures alerted the team first. Every change to the 18 stores went through review and a deploy I owned, and the ERP/POS integrations became one of the main workstreams of the Shopify migration I led.