Practical AI systems, built around your work.
I help businesses automate repetitive work and connect AI to the systems they already use. If your team is already building, I can take on a scoped piece of implementation or testing.
What I do
The starting point is the work: what comes in, who handles it, where it gets stuck, and what a useful result looks like. That might lead to a document workflow, a connection between systems, or a small internal tool.
We check what your existing software can do before building something new. Sometimes a configuration change or a simple script is enough; AI has to earn its place.
Common projects
- Turn scattered email, forms, and PDFs into routed workflows. Less time copy-pasting between systems, fewer dropped handoffs.
- Wire AI into the tools you already use: CRM, docs, email, support, spreadsheets, internal databases. Your team keeps their workflow; the AI does the busywork.
- Build internal agents and automations with audit trails and human review. Define approval steps, record what happened, and test how the workflow handles failures.
- Replace manual reporting, research, and admin loops with small reliable systems. Compare time spent and error rates before and after the change.
- Take on a defined item from your technical backlog. Integrations, evaluation tooling, and workflow components, with documented dependencies and a handoff to your team.
Already building with AI?
Your team may know what needs building but have more work than time. I can take on a defined integration, prototype, or evaluation task within your architecture. Your technical lead sets the direction; we agree on interfaces, access, and acceptance checks before I start.
The work includes tests, documented dependencies, and a handoff your team can use. Code ownership and any vendor or licensing constraints are made explicit in the scope.
Who I work with
Business owners, operations leads, and internal technical teams in Michigan and Texas. I'm based in Northern Michigan, in Central Texas regularly, and most work happens remotely.
You might be choosing a first workflow to automate or already have a working system and a backlog. Either way, we start with a specific problem and the constraints around it.
How I reduce delivery risk
Before implementation, we agree on acceptance checks using representative examples, including missing data, incorrect outputs, and service failures. We define which data the system can access and what it should do when it cannot produce a usable result.
Your team decides what requires human approval and who can change those rules. Logs and test results inform that decision; they don't make it. If the goal is saving time or reducing errors, we capture a baseline and compare the results.
Personal software case study: testing and deployment in DocLifts →
How engagements work
We start with a 20-minute call about the task, your existing systems, and your constraints. If there's a fit, the proposal defines the deliverable, acceptance checks, dependencies, and price. Delivery includes a documented handoff and an agreement on who operates it afterward.
Pricing: $5k fixed-price floor on scoped projects, $250/hr hourly when a project doesn't fit a clean scope. Flex on scope, not on rate.
Bring a workflow, a backlog item, or a technical question. We'll work out whether there's a useful piece of work I can take on.
Book a 20-minute call →