Docker Documentation Gets an AI-powered Assistant | Docker

We recently launched a new tool to enhance Docker documentation: an AI-powered documentation assistant incorporating kapa.ai. Docker Docs AI is designed to get you the information you need by providing instant, accurate answers to your Docker-related questions directly within our documentation pages.

Docker Docs AI

Docker documentation caters to a diverse range of users, from beginner users eager to learn the basics to advanced users keen on exploring Docker’s new functionalities and CLI options (Figure 1).

Navigating a large documentation website can be daunting, especially when you’re in a hurry to solve specific issues or implement new features. Context-switching, trying to locate the right information, and piecing together information from different sections are all examples of pain points users face when looking up a complex command or configuration file. 

The AI assistant addresses these pain points by simplifying the search process, interpreting your questions, and guiding you to the precise information you need when you need it (Figure 2).

Find what you’re looking for

Docker documentation consists of more than 1,000 pages of content covering various topics, products, and services. The docs get about 13 million views every month, and most of those views originate from search engines. Although search engines are great, it isn’t always easy to conjure the right keywords together to get the result you’re looking for. That’s where we think that an AI-powered search can help:

• It’s better at recognizing your intent and personalizing the results.
• It lets you search in a more conversational style.

More importantly, kapa.ai is a Retrieval-Augmented Generation (RAG) system that uses the Docker technical documentation as a knowledge source for answering questions. This makes it capable of handling highly specific questions, contextual to Docker, with high accuracy, and with backlinks to the relevant content for additional reading.

Language options

Additionally, the new docs AI search can answer user questions in your preferred language. For example, when a user asks a question about Docker in Simplified Chinese, the AI search detects the language of the query, processes the question to understand the context and intent, and then translates the response into Simplified Chinese (Figure 3). 

This multilingual capability allows users to interact with the AI search seamlessly in their native language, thereby improving accessibility and enhancing the overall user experience.

Using the Docker Docs AI

We’re thrilled to see that our users are highly engaged with the AI search since its launch, and we’re processing around 1,000 queries per day! Users can vote on answers and optionally leave comments, which provides us with great insights into the types of questions asked and allows us to improve responses.

The following section shows interesting ways that people are using Docker Docs AI.

Answers from multiple sources

Sometimes, the answer you need requires digging into multiple pages, extracting information from each page, and piecing it together. In the following example, the user instructs the agent to generate an inline Dockerfile in a Compose file. 

This specific example doesn’t exist in the Docker documentation, but the AI assistant generates a file using different sources (Figure 4):

In this case, the AI derived the answer from the following sources:

Debugging commands

Often, you need to consult the documentation when you’re faced with a specific problem in building or running your application. Docker docs cannot cover every possible error case for every type of application, so finding the right information to debug your problem can be time-consuming. 

The AI assistant comes in handy here as a debugging tool (Figure 5):

Here, the question contains a specific error message of a failed build. Given the error message, the AI can deduce the problematic line of code in the Dockerfile that caused this error, and suggest ways to solve it, including links to the relevant documentation for additional reading.

Contextual help

One of the most important capabilities unlocked with AI search is the ability to provide contextual help for your application and source code. The conversational user interface lets you provide additional context to your questions that just isn’t possible with a traditional search tool (Figure 6):

Dive into Docker documentation

The new AI search capability within Docker documentation has emerged as an indispensable resource. The tool streamlines access to essential information to a wide range of users, ensuring a smoother developer experience. 

We invite you to try it out, use it to debug your Dockerfiles, Compose files, and docker run commands, and let us know what you think by leaving a comment using the feedback feature in the AI widget.

Explore new Docker concept guides

• What is a container? This guide includes a video, explanation, and hands-on module so you can learn all about the basics of building with Docker. 
• Building images: Get started with the guide for understanding the image layers.
• Running containers: Learn about publishing and exposing ports.
• GenAI video transcription and chat: Our new GenAI guide presents a project on video transcription and analysis using a set of technologies related to the GenAI Stack.
• Administration overview: Administrators can manage companies and organizations using Docker Hub or the Docker Admin Console. Check out the administration manual to learn the right setup for your organization.
• Data science with JupyterLab: A new use-case guide explains how to use Docker and JupyterLab to create and run reproducible data science environments.

Learn more