We build automation for small businesses that need it to actually work.

Headwaters Automation is led by Tom Faries, based in Portland, Oregon. We help small and mid-sized businesses replace the slow, manual parts of their operations with AI and automation systems that run reliably, so the team you hired can spend their time on the work you hired them for.

Before founding Headwaters, Tom spent more than sixteen years across LinkedIn/Microsoft and growth-stage startups like Metadata and Pantheon, leading and selling into brand marketing, performance marketing, marketing operations, revenue operations, and sales. At Pantheon he worked closely with developers and the IT infrastructure leaders who keep modern web platforms running. Across all of it the same pattern kept showing up, both inside the companies and across the businesses they sold into: smart people stuck running workflows that should have been running themselves, painful manual processes everywhere from HR to legal to sales to marketing, and tools that promised to fix it but never quite did. That is who Headwaters was built for. Businesses that deserve operations that hold up, without paying for a six-figure consultancy or a full-time ops hire.

The name comes from the headwaters of a river, the point where everything downstream starts. That is where we like to work in a business. The upstream systems (the ones that produce leads, serve customers, and generate reports) set the quality of everything that flows from them. Get those right and the rest of the organization runs cleaner.

Our client roster is intentionally small. Small enough that every engagement gets senior attention. We take fixed-scope projects with written success criteria. If we cannot tell you at the start what a system will do and how you will know it is working, we will not build it. Every engagement draws on a senior practitioner network, with the right specialists matched to the scope.

How we think about AI

Most AI consulting right now is selling possibility. We are more interested in what ships and what it actually returns. That means we prefer the unglamorous pattern that recovers ten hours a week, every week, over the impressive demo that no one runs after launch. It also means we will tell you when a workflow is better handled without AI at all, because the goal is time back, lower costs, and fewer dropped balls, not a story about how much AI you used.

AI is one tool in the automation stack, not the whole stack. A good automation team knows when to reach for it and when to reach for something simpler.

What we care about

Systems that keep running on their own, day after day. Access and logins managed so nothing quietly breaks a year from now. A clear record of what the system did and when, ready when you need to look back. Safety checks anywhere it touches a customer. Documentation written in plain language that still makes sense six months later, to a real person, not an engineer.

If that sounds like what you want, start here.