AI Strategy & Creative Workflows
Find where AI
can help.
Make room for useful assistance—and keep judgment with the people doing the work.
I help small businesses and cultural, educational, and nonprofit organizations explore practical ways to research, draft, and organize material with AI. We can review a workflow, test a small pilot, or work through the tools together.
Discuss an AI workflow ↗01 / What we can explore
One useful task.
A process you can assess.
You may be experimenting with AI without a shared approach, or wondering whether a repetitive task is worth changing. We begin with your work and constraints, then choose a manageable question to investigate.
Workflow reviews
Map how work moves from request to delivery, including source material, decisions, and the checks people already make.
Possible deliverables: a workflow map, opportunities to investigate, and clear points for human review.
Small pilots
Test a bounded AI-assisted task against agreed criteria before deciding whether to use it more widely.
Possible deliverables: a pilot brief, test examples, a documented process, and a short findings report.
Practical guidance
Work through a task together, refine your instructions, and learn how to assess and edit the output.
Possible deliverables: reusable instructions, review criteria, and notes your team can return to.
XYZ Company / Inquiry workshop
A thoughtful first response.
Guided simulation · Fictional example
01 / Project details
Living room, dining room, and two bedrooms. Walls reported in good condition. Requested timing: next month.
02 / Review decision
Prepare a consultation discussion.
A representative reviews the details before making commitments.
03 / DRAFT ONLY · NEVER SENT
Pricing, availability, and completion dates require review by a representative.
02 / See the thinking in practice
Organize the inquiry.
Keep a person in charge.
The XYZ Company inquiry demonstration shows how project details could become a summary, a review decision, and a draft response. It makes the handoff to human judgment visible.
This is a guided simulation with five fictional scenarios and prepared results. It does not run live AI, accept new inquiries, or send messages. No time savings or business results are claimed.
Explore the guided inquiry demonstration ↗
Read the full campaign case study →03 / How we’ll work
A clear task.
Room for human judgment.
- 01
Define the experiment
We clarify the goal, suitable inputs, constraints, tool costs, and what a useful result would look like. We address confidential material and access before testing.
- 02
Test & review
We try representative examples, check facts and tone, and record errors, editing effort, and human corrections. Where possible, we compare with the existing process.
- 03
Decide what follows
We document the process, its limits, and the review responsibilities. The recommendation may be to continue, revise the approach, or stop.
A focused starting point
Start with one content task.
A possible pilot could turn an approved organizational report into a draft newsletter. We would assess factual accuracy, voice, editing effort, and whether the process is useful.
An illustrative engagement, not a completed client project.
Know what’s included.
We agree on the task, inputs, deliverables, fee, timing, and review criteria. Someone owns the outcome and makes the final decision.
Custom integrations, autonomous multi-step systems, paid tool subscriptions, and ongoing support require a separate scope. AI output is reviewed before use.
What part of your work
could use a closer look?
Bring a recurring task, an idea, or a question about the tools. We’ll identify a manageable first experiment and agree on the work before testing begins.
Let’s discuss your workflow →