A Closer Look At API Docs Generated via LiveAPI's AI

I've been working on an engineering tool, LiveAPI, for the past few months, with a passionate team of 5.

We are mainly focused on automating the documentation generation process.

We think too many teams spend too much time putting together a few API reference generation pipelines, or more often, they skip the work altogether - because it is too much effort.

With LiveAPI - our goal is to give engineers a more convenient yet highly effective solution.

Our pact is simple: You connect your git repos, we give you an always up-to-date documentation link for your repository.

On top of that - we provide many valuable features on top of this basic layer - that helps you:

1. Discover APIs in an organized way
2. Get guidance on how to use those APIs
3. Get code auto-generated - which you can copy/paste into frontends, etc
4. Search for APIs (within a project as of now. Global search is planned)
5. Mechanisms to maximize privacy while obtaining great convenience

Our goal is to perfect the product, but as is usually the case - we are still a work in progress.

I'd like to quickly show an example document produced for a service called fb_backends:

This whole documentation was automatically generated on a new commit - within 2-3 minutes.

Notice these aspects in the screenshot:

1. Organized APIs. Many categories have been automatically detected
2. "Try It" Editor. Built on our Open Source Lama2 initiative, this editor lets you try every API right within the browser.
3. API Descriptions
4. Search Box

Also, you can see how I've generated Javascript code out of the request with 1 click:

Here's how search looks now:

This is functional at a basic level, but we do recognize many improvements that can be done:

1. Better visuals; more customized visuals as well
2. Automatic SDK generation & downloads
3. Input/Output data structure specification
4. Support images in the "Try It" editor correctly
5. Advanced authentication support
6. Links to source code locations of relevance
7. Better differentiation between API method types
8. Pre-filled example outputs, better sample input variables

We want to take on this challenge in the upcoming weeks, and 10x the customer convenience.

We'd love to hear your feedback

One thing we love at Hexmos is Customer Feedback. We are eager to listen and improve the product. So here's a closing question for you:

• How can we improve this API documentation - based on your experience dealing with APIs and API documentation?

Do let us know in the comments.

Learn More

You can Learn More about LiveAPI at:

1. LiveAPI Home
2. How LiveAPI can be a strong Swagger Alternative
3. How LiveAPI can act as an automated alternative to Redocly