- Home
- Services
- AI systems
AI systems
Show me the task your team keeps doing by hand.
I'll tell you whether AI belongs in it. For one property manager, I built tooling that prepares each new listing as a paused ad draft. A person checks it before it can run.
Start with the work, then pick the tool.
Start with one task your team already knows how to do. Write down the steps and judgment calls, test them on real examples, then decide whether the task should run on its own.
Task
You do it by hand, once, and notice every step and every judgment call you make.
Job description
The steps written down: who the AI is, who it writes to, how you sound, what good looks like. Paste it in. Now it repeats.
Agent
The job description gets real inputs on its own: your inbox, your calendar, a form. It runs, you review.
Automation
On a schedule or a trigger. It runs without you in the loop, and you check the output.
What changes
Let the tool prepare it. Keep the decision with your team.
We do the same setup work every time a new job comes in.
The work
For one property management client that meant tooling that creates each listing ad paused and review ready, so what used to be a blank screen is now an approval. For a paid media account it means the weekly report writes itself from the platform data and waits in drafts for a read.
I build around the steps your team already uses, including where a person needs to check the result.
Human outcome
A new listing produces an ad draft. A person checks it before the ad can run.
Common failures
Why the first AI draft usually needs work
Generic copy
What changes
Write down your voice, banned phrases, and rules for numbers. Check each draft against them before it goes out.
Missing context
What changes
Give the workflow the client and project facts it needs, then check that it used the right version.
No review step
What changes
Work that spends, sends, or publishes waits for approval. You see the proposed action and its dollar impact first.
Three layers of support
I can help you choose, build, or maintain the system.
-
AI consulting
Find the opportunities worth doing, and name the ones that are not. A roadmap that says which task climbs which rung, in what order, and what it is worth.
-
Custom builds
Build the thing itself: an automation, a dashboard, an intake or reporting workflow. Fixed scope, fixed price, an end date, and an approval gate on anything that leaves the building.
-
System care
Keep it working when the process changes, because it will. Monthly, with a log of what ran, what it proposed, and what you approved.
-
Workshops
The same method taught in a room: pick the one thing that ate your week and hand the boring half to AI without losing your voice. See the sessions.
Start here
The AI Clarity Session. $1,500, one week, a written roadmap.
One three-hour working session on your real tasks, then a written roadmap inside a week: which task climbs which rung, what it costs, and what it is worth. You keep it whether or not you build with me.
- Before
You pick the one thing that ate your week
A short intake: the task, who does it, how long it takes, where the judgment calls are.
- Session
Three hours, your tools, your examples
We do the task by hand once, write the job description, and test it live. You leave with rung one working.
- Within a week
The roadmap
Every candidate task, its rung, its build cost, and its hours back. Priced honestly, including the ones not worth doing.
AI systems and automation
Tell me which task keeps coming back.
Describe the repeat work and the tools involved. I'll tell you what I would check first and whether an automation makes sense.
Prefer to talk it through? Book a call. The AI Clarity Session is there if you want to work through one task together.