One day I tried out Cursor with GitHub Copilot and the Copilot cli. I saw ai agents are really here and ready to build with humans doing real software engineering work. I built more and learned more becoming an agentic engineer and builder. Now, I build agentic layers into codebases so they can work on themselves and learn how to build themselves better.

Growing businesses need growing systems. Gaia grew quickly. More customers watching and reading more content in more languages on more devices. The engineering team migrated a vendor-hosted monolith to containerized microservices running in a global hybrid cloud. We started streaming live events from an on-site event center. Along the way, I explored how AI could integrate into content management and engineering workflows.

Small and medium businesses needed real tools, not enterprise one-size-fits-all subscriptions. I led global distributed teams building custom open source solutions and supporting cloud migrations. We overhauled legacy systems to take advantage of the services available through AWS.

I led a small team modernizing legacy systems that were holding the organization back. We moved everything to AWS and rebuilt our payment processing systems to eliminate PCI compliance headaches and risk. The cloud-based systems were a huge cost savings over the previous stack of servers in the data center.

After building and updating so many sites on so many platforms, I thought I should host my clients’ sites. So I hosted and managed PHP, Drupal, and WordPress sites for clients who had better things to do than fight with servers.

The focus was on clean design and simple, elegant functional. I crafted websites with HTML, CSS, JavaScript, PHP, and Perl back when these tools were still finding their place.

The commercial internet was still figuring itself out. Two friends and I started a web design company in our dorm room at CU. We built custom sites for clients who were brave enough to invest in the web before it was obvious that they should.

Designed and built websites in the early web days. It was all new then, and every project was figuring out what the web could be.