Restructuring AI chatbot documentation for user adoption

Technical Writer · 2025 · 6 months · 3 people · 3 min read

Transformed a lengthy chatbot user guide into structured, task-oriented docs and created a prompt library to drive adoption.

Overview

Redesigned Vertex's AI chatbot user guide from a single long-form topic into a persona-based structure with organized sections and a curated prompt library. The restructuring made the documentation more navigable and helped users quickly understand what the chatbot could do.

Problem

The original user guide was a single, longform topic covering management, usage, and best practices for the chatbot. New users found it overwhelming and hard to locate task-specific information. There was no clear way to showcase the chatbot's capabilities or help users get started quickly.

Approach

I restructured the guide into digestible, task-focused topics organized by user personas. I also built a prompt library that contains a collection of example prompts and responses organized by use case. The prompt library provides users quick access to what they need and visibility into what the chatbot can do.

Challenges

  • Restructuring had to accommodate existing content (couldn't start from scratch)
  • Upcoming product features were in flight, requiring flexible information architecture
  • No usage metrics to measure impact, only stakeholder feedback

Key Tasks

Build a prompt library

Reasoning:

Users didn't need to read the full guide to see what the chatbot could do. A curated library of working prompts showed capabilities immediately. I collaborated with the product manager to identify the use cases that would resonate most with our users. This made the library focused and relevant instead of overwhelming.

Alternatives considered:
  • Scatter examples throughout the guide (harder to find and reference)
  • Omit examples (users wouldn't know the types of questions they can ask)

Organize restructured content by task, not feature

Reasoning:

Breaking the longform guide into task-focused chunks (how to manage, how to build prompts) made it scannable. Users find what they need in minutes instead of scrolling.

Tech Stack

  • AuthorIt

Impact

  • Restructured from 1 long document into scannable, task-focused sections
    Documentation structure
  • Created prompt library with curated prompts organized by use case
    Feature showcase
  • Users can complete their tasks in minutes instead of scrolling through content
    User efficiency

Stakeholders reported the restructured guide was significantly easier to navigate. The prompt library also allowed new users to see the chatbot's capabilities immediately and start creating their own prompts. The task-focused structure also made maintenance straightforward when new features shipped.

Learnings

  • Topic-based authoring makes content more digestible and reduces cognitive load for users
  • Example libraries can be adoption drivers
  • Restructuring existing content is often more valuable than writing from scratch

Notes

The most important thing I did for this project was build the prompt library. Working with the product manager to identify high-value use cases meant the library showcased what actually mattered to users. And as new features ship, we can add or update example prompts easily.

This project also say this taught me that working examples are the quickest way to showcase a product’s value.

All projects