One-on-one AI mentoring

Turn your expertiseinto AI workflows thatget better over time.

Private mentoring for professionals who want more than prompts. Learn how to translate judgment, standards, and context into repeatable systems.

You will learn where AI helps, where it needs oversight, and how to keep improving the workflow after each run.

Work with meSee how it works
Mentoring on your work

Practical guidance on real tasks.

Repeatable workflows

Capture your thinking in a process.

Results you can trust

Match review to the stakes.

Client testimonials

What clients say about working with Tom

Dan Lerman
Tom is the perfect teacher for this. He is a masterful tutor, meeting you at your level and mindfully guiding you to grow.

Dan Lerman

Columbia Professor | Advantage Testing

1 / 3

I use AI sometimes, but I do not know where it belongs in my week.

You have seen what AI can do. You want to identify the tasks where it can help often enough to become part of your routine.

A useful first task.

I have seen AI produce strong work. I want that quality on my own tasks.

You want to learn which context, examples, and instructions consistently lead to a result you would use.

A repeatable recipe.

I need AI results I can depend on.

You want direct checks for facts, stronger safeguards for consequential actions, and lighter supervision when the risk is low.

Checks that match the stakes.

New tools keep arriving. I want to know which ones deserve my attention.

You want to judge each tool by the work it can do for you, using your own tasks as the test.

A better way to choose tools.

What you learn to do

Build one dependable AI routine, then reuse the method on the next job.

You learn how to choose the task, supply the right context, and test the result at the right level. Those skills stay useful when tools change.

Save time on a job you already do

Start with a specific task such as meeting prep, first drafts, research summaries, health tracking, or project updates. The task should happen often enough for a better routine to matter.

Each useful correction improves the next run

When a result misses, find the instruction, example, source, or check that would have prevented it. Save that correction so later runs start stronger.

One working routine gives you a method for the next

Once you can build and adjust one useful routine, you can apply the same method to another task. The method stays useful even when the tools change.

Featured book

Becoming an AI Power User

A principle-first guide to building an AI practice that compounds. Power users build compounding loops: they experiment at the frontier, reject weak output, diagnose failures, meta-prompt improvements, save reusable assets, evaluate with clear standards, and retest as models change. This book teaches the durable principles through practical projects: quality ratcheting, meta-prompting, error-to-spec transformation, context engineering, experimentation as R&D, evaluation loops, separation of concerns, source verification, portable infrastructure, creative building, judgment preservation, and workflow redesign.

524 min read38 chaptersAIPower UserCompounding Loop

Read the book
A deep blue Becoming an AI Power User hardcover beside context, iteration, evaluation, and saved-work cards.
  • Keep related ideas connected
  • Leave each chapter with a next step
Tom Meli

About Tom

Since 2023, I have coached people across dozens of fields.

We start with one useful job, test the routine on their own material, and save the instructions or checks that make later runs better. That pattern keeps holding.

One-on-one AI mentoring

Get more donein the hours you have.

Bring the knowledge work that keeps getting stuck, scattered, or repeated — research, drafting, planning, follow-up, or decision prep. In one working session, we build AI workflows on your real material. You’ll leave with usable workflows, clear next steps, and a method you can extend.

Tell me what you're working on

One sessionYour real workBuilt live with you