🏛️Diataxis Framework

The Diátaxis framework is a structured approach to organizing documentation, designed to make it easier for users to find the information they need based on their goals and the specific context in which they’re using a product. Diátaxis was created to address the common problem of documentation overload by categorizing content into clear, user-focused sections, each tailored to a specific purpose.

The Four Quadrants of Diátaxis

Diátaxis divides documentation into four main types, each serving a unique purpose:

1. Tutorials – Learning by Doing Tutorials guide users through practical steps to achieve specific outcomes. They’re hands-on and goal-oriented, teaching users by walking them through a task or process. Tutorials are ideal for beginners or users looking to understand how to accomplish something concrete.

2. How-To Guides – Focused Problem Solving These are short, task-specific guides that provide steps for accomplishing specific tasks or solving problems. How-To guides are helpful for users who have some familiarity with the product and need specific solutions quickly. These guides focus on “how” to do something efficiently without going into extensive theory.

3. Reference – In-Depth Information Reference materials are comprehensive and cover the product’s features and functionalities in detail. They act as the go-to resource for precise information, including technical details, API endpoints, and configurations. Unlike tutorials or guides, reference documentation is not task-based; it’s structured for users to look up specific information as needed.

4. Explanations – Deep Dives into Concepts Explanations clarify the “why” behind features, design choices, or approaches. They provide context and background information, helping users understand the underlying concepts and rationale. Explanations are particularly useful for users who want to grasp the big picture or make more informed decisions about using the product.

Diátaxis in the Context of Katara AI

In Katara, the Diátaxis framework can streamline user interactions and learning by clearly distinguishing each type of documentation:

• Tutorials could cover step-by-step setups, like creating an agent from scratch, configuring it for a specific task, or integrating Katara AI with other tools.

• How-To Guides would focus on common workflows within Katara, such as using RAGs for retrieval, managing corpuses, or fine-tuning agents for specific purposes. These guides allow users to quickly find solutions to specific questions.

• Reference Documentation would include technical specifications for each feature, like the available base LLM models, detailed configuration options for agents, or API calls. This is crucial for developers or advanced users who need in-depth information.

• Explanations would explore the philosophy behind Katara’s approach to AI and workflows, explaining how components like RAGs or base models enhance accuracy and performance. These sections help users understand why certain decisions or setups are beneficial, leading to a more cohesive understanding of Katara AI.