I Build AI Systems That Businesses Own

I spent years in ecommerce, mostly on Amazon. I scaled a brand to $400k in its first year and went on to become president of an Amazon SaaS company. That job taught me how subscription software is priced and packaged from the inside, and how much of what customers pay for they never use.

When AI search arrived, I wanted real data on how it worked, so I built the infrastructure to get it: browser automation across four AI platforms, a pipeline that collected 100,411 citation events, a hybrid vector and graph knowledge base, and an analytics product that replaced a paid SaaS category. That work became six DOI-archived research papers and two open-source projects.

Along the way I realized the systems were more valuable than the subject. The same pipelines, knowledge graphs, and data connectors solve problems every small business has: too much manual work, knowledge nobody can find, and software bills that only go up. That is what AI+Automation builds now.

AI should amplify people, not replace them. The best automation removes the work your team hates so they can do the work only they can do.

Boring and reliable beats clever. I keep the control flow in plain code and use AI only where it adds judgment. Systems should do the same thing every time and tell you when they cannot.

You should own what you pay for. Code, data, and accounts belong to you. If we stop working together, everything keeps running.

Honest scoping. If an off-the-shelf tool solves your problem for $30 a month, I will tell you that on the first call.