Skip to main content

Ninja Guide 🥷 Documentation to Enhance User Experience 📕

Learn how to build your document library to power the Expertise feature and Open Learning.

Updated today

Documentation is the centralized document library of your muchbetter.ai environment. It holds your official documents (product sheets, pricing, terms and conditions, sales guides…) and powers two key platform features.


🎯 What is Documentation used for?

Feature

What it does

🔍 Expertise

Checks the factual accuracy of what the learner says during a simulation: pricing, guarantees, conditions, product features… by cross-referencing with your documents.

📚 Open Learning

Lets learners ask questions to the AI assistant via the Knowledge tab (Practice space). The AI answers based on your documents without displaying them.

💡 Expertise checks the substance (correct information). Evaluation criteria check the form (correct behavior). They complement each other.


📄 Adding a document

Go to Admin → Documentation → "Add a document".

  • Accepted formats: PDF, DOCX, PPTX.

  • Naming: this is critical. The document title must be precise and descriptive. E.g. "Pro Insurance Plan: 2025 Pricing". The AI uses this title to find the right information.

⚠️ A vague or mismatched title means the AI may not find the correct information.

💡 "AI-friendly" documents: use structured text, clear headings, and clean paragraphs. Avoid complex tables, graphic layouts, multi-column designs, and scans. If a document is heavily formatted, create a simplified version alongside it.


🔗 Activating Expertise on a situation

Once your documents are uploaded, you can link them to a situation to enable Expertise:

  1. Go to Admin → Situations → your situation → Configuration tab.

  2. Check "Enable Expertise".

  3. Enter the exact titles of the relevant documents or sections. The more precise, the better the AI will target its search.

When a factual error is detected during a simulation, Expertise displays:

  • The exact passage from the transcript containing the error

  • The corresponding audio

  • What should have been said


🔐 Security and access

What the learner CAN do

What the learner can NEVER do

Ask the AI assistant questions via Knowledge (AI answers without showing the docs)

Access the document library list

View documents attached to a persona during a simulation (if enabled)

Download or export a PDF

See information from a document not enabled for their situation

💡 Your proprietary documents (pricing, terms, internal guides) are secured within the environment. Only the AI can access them to answer questions and check factual accuracy.


📖 Vocabulary: helping the AI recognize your business terms

🧐 Where? Settings → Vocabulary

Vocabulary is a companion tool to Documentation. It teaches the AI to correctly recognize and pronounce words specific to your business, brand, or industry: product names, acronyms, technical jargon, foreign terms…

Without Vocabulary, the AI may mishear what the learner says (speech recognition) or mispronounce a word during the simulation (text-to-speech). Vocabulary fixes both.

➕ Adding a word to Vocabulary

In Settings → Vocabulary → "Add", fill in 3 fields:

Field

Required

Detail

Vocabulary word

Yes

The exact term as it should be recognized and written. E.g. "Wifi"

Description

Optional

A short definition to help the AI contextualize the word. E.g. "Wireless network"

Pronunciation

Optional

Phonetic spelling to guide text-to-speech. E.g. "wy-fy"

💡 When to use it:

  • Product or brand names that aren't standard words

  • Internal acronyms (e.g. GDPR, CRM-specific terms…)

  • English terms in a French-speaking environment (or vice versa)

  • Industry-specific jargon

  • Any word the AI consistently misspells or mispronounces

💡 Documentation + Vocabulary = winning combo. Documentation gives the AI knowledge (what to say), Vocabulary gives it words (how to say and recognize them).

Did this answer your question?