Wellness workspace
An opinionated, white-label AI wellness coach built on Paul Chek’s teachings. Clients work through lessons and coaching conversations in their own private workspace and decide what their practitioner sees. The practitioner gets an AI brief of what’s been shared, with every point tied to the client’s own words.
I built it for my own coaching practice and for other CHEK practitioners, and each practice runs it under its own name. My own practice is set up in it, and so is Franki Wellness, the practice of Franki King, whose website I rebuilt. The screenshots follow a made-up client, Nora Lind, her practitioner, Sam Okafor, and their practice, Riverbend Wellness.
Coaching grounded in his approach
The coaching covers five topics: direction and happiness, stress and capacity, rest and rhythm, food and water, and movement that fits. Those topics and the lessons are my own adaptations of Paul Chek’s material, and each lesson shows which of his courses or books it draws on. Coaching starts with a 136-question intake, and from then on the coach works from those topics and from what the client has chosen to save. The app asks before it sends anything to the AI.
Private until the client chooses
Lessons and conversations shared with a practice are the newest part, now in a private preview. Each lesson ends with a few reflection questions, and the answers stay in the client’s own account. When Nora wants to talk one through, she brings that single answer into a new coaching conversation. Before her first message, she chooses who can read it: just her, or her practice. She can type that message, or speak it and edit the transcribed draft before sending. A conversation she keeps with her practice stays apart from her personal conversations and records.
What her practitioner sees
Sam works in a separate practitioner workspace and sees only what his clients share with the practice. Clients can already send him a guidance request, a question along with observations they choose, and he replies.
In the preview, he can also read the conversations clients kept with the practice. So he doesn’t have to read every message, the practice’s AI assistant prepares a brief. Every point in it sits above the client’s exact words and links to where they said them, so Sam can check the AI’s reading against what Nora actually wrote.
Each practice under its own name
White-label here means each practice brings its own name, welcome message, website and color into the same app. That part is already live: clients see the practice’s identity when they’re invited and after they connect, and they can end the connection whenever they want.
From coaching to practice
Coaching can lead to practice too. The workspace has a career-change topic in development, which answers with page citations from a private library of career-coaching guides. From one of its replies, a client whose account I’ve connected can rehearse an interview in the practice tool on my personal server, and the result comes back into the same conversation. If they want their practitioner’s view, they choose one result to share.
My own study works differently. Practice that starts from a book in my library, like Pitch Anything, returns to that book, and only practice that starts in a coaching conversation can reach a practitioner, one chosen result at a time.
On the web, on iPhone and in ChatGPT
All three use the same account, so my own record opens in each. The web app is live and is where practices are managed. The iPhone app covers the personal workspace and is a private beta on my phone, and the ChatGPT plugin is installed only in my own ChatGPT account.
Where it stands
The web app is live with personal and practice workspaces, the intake, the five coaching topics, guidance requests and practice branding. Lessons, conversations kept with a practice, spoken messages, the AI brief and the career-change topic are in a private preview. If you run a practice, or you’re building something similar, I’d like to hear from you.