> ## Knowledge Base Index
> Fetch the complete knowledge base index at: https://help.klarity.coach/sitemap.xml
> Use this file to discover available pages before exploring further.
> Pure-Markdown content can be obtained by appending a '.md' suffix to the content URLs listed in the sitemap (without the trailing slash).

# Personalizing the AI with your practice

The better the AI knows how you coach, the more its texts sound like you. The **AI personalization** page gathers the — optional — details Klarity uses to tailor your briefings, Memory Chat and end-of-engagement booklets to your practice.

# Understanding AI personalization

Open **Settings › AI personalization** (left-hand Settings menu). The **Your practice (AI)** page groups information about the way you work. It only feeds AI-generated texts: it is **never shown to your clients**.

| Every field is optional: the AI works without them, but each detail refines the tone and angle of its writing. Changes can take **a few minutes** before they are reflected in AI answers.

# Describing your practice

1. **Approaches & methodologies** — the frameworks that structure your questions and interventions. Add them with the selector (18 approaches offered: solution-focused, cognitive-behavioral, systemic, Socratic questioning, NLP, transactional analysis, positive psychology, narrative approach, somatic approaches, directive or non-directive stance, Gestalt, co-development, appreciative inquiry, mindfulness, Ericksonian hypnosis, Process Communication, neurocognitive approach).
2. **Audiences** — who your coaching most often serves (15 audiences offered: executives, managers, entrepreneurs, employees, professionals in transition, young professionals, students, private individuals, teams and organizations, women in leadership, atypical profiles, healthcare professionals, athletes, end-of-career, teenagers).
3. **Years of experience** — a range, from "less than 2 years" to "more than 20 years".
4. **My practice** — free text (1,000 characters max, with a counter) where you describe in your own words how you work, your tools, what you pay attention to. It is the richest field for the AI.
5. Confirm with **Save changes** in the bar that appears at the bottom of the page.

| This is the same information as the "About you" step of onboarding: you can complete or correct it here at any time.

# The methodology lens

Under "Your practice", the **Methodology lens** card has the AI read your pre-session briefings and session summaries **through one or two of your coaching frameworks**.

1. **Reading frames** — choose up to **2 frames** among your declared methodologies (**Add a frame** button). By default, your first two declared methodologies are used. With no active frame, the methodology lens is **off**.
2. **Apply to pre-session briefings** — adds a **"Preparation angles"** section to the briefing, read through your frames.
3. **Apply to session summaries** — adds a reading section to the summary (for example **"Reading — Transactional Analysis"**), flagged with the **Framework reading** badge.

| This reading is for you only — it is **never included in the client email or the end-of-journey booklet**, and applies to **one-to-one sessions** only.

# The "Your AI profile completeness" gauge

At the top of the page, a gauge shows how much of your AI profile is already filled in. It counts **7 items**, listed under the bar as chips: filled items are checked, items **still to fill in** appear dashed.

- **On this page** — the approaches, the audiences, the experience and "My practice".
- **On the Public profile page** — the **headline**, the **bio** and the **specialties**. A note under the gauge points you there: these three items are set in **Settings › Public profile**.

| The gauge is a guide, not an obligation: a partially filled profile already improves the AI's writing.

# The observed style (auto-calibration)

On top of what you declare, the AI can **observe your style** in your session summaries and propose a short description for you to validate.

1. Turn on the **Observed style (auto-calibration)** switch. Each week, the AI can then infer a style proposal from your summaries.
2. When a proposal is ready, a **New observed style to review** box appears, with the proposed text and its observation date.
3. You have three choices: **Approve** (the style is used in your AI actions), **Edit** (adjust the text before approving) or **Reject** (the proposal is dismissed — the AI will suggest a new one within a month).

An approved style stays displayed with the caption **Approved on {date} — used in your AI actions**. As long as no proposal exists, the page notes that the observed style will be suggested **after a few summarized sessions**.

## When your style evolves

Once a style is approved, the AI keeps observing it across your sessions. When it spots an evolution, a **Your observed style has evolved** box shows your **Current style** and the proposed **Refined version** side by side (with its refinement date). You can **Approve** the refined version (after **editing** it if needed), or **Reject** it: nothing changes until you decide — declining keeps your current style, and the page says so (**Refined version dismissed on {date} — the AI may suggest a new one within a month**).

|| Nothing is applied without your agreement: a style proposal only influences the AI **once you approve it**. Turning the switch off pauses new proposals.

| 💡 **Tip:** invest in **My practice** above all: two or three concrete sentences ("I work through questioning, I use Ikigai…") do more for briefing quality than any number of checked boxes.

# Related articles
- Completing your coach profile
- The pre-session briefing
- What is the Memory Chat?
- The end-of-engagement booklet