ReadmeAI

Inspiration

• The success behind every open source project is its Readme.
• The developer doesn't spend much time on writing documentation or readme.
• With ReadMeAI we solve a major pain of any developer and that is writing project Documentation.

What it does

• At first, the user has to provide a Project source code file
• And have to write a short description of the project in 15 to 65 words.
• After clicking on generate it may take 7 to 15 seconds to generate readme (the generation time depends on the length of the project)
• After generation, the user can see its readme markdown and its preview
• We have also provided a built-in editor to modify the markdown, with this editor any developer without any markdown knowledge can write a markdown.
• At the end, the user can download the readme file in markdown and html format.

How we built it

• We have used javascript programming language for our project
• We have built the front end of the site with core technologies like HTML, CSS, and JavaScript.
• At the backend, we have used nodejs runtime and express js framework to write business logic and REST APIs
• For content Generation, we are using Google PaLM AI, which is free to use.
• We're also using prompt engineering techniques to generate a specific output from any given user input.
• We have also containerized the ReadMeAI project using Docker for easy deployment.

Challenges we ran into

• The GitHub collaboration workflow is a challenge for us, we learn about how we can communicate as a team and how we can share and implement our ideas or features.
• Another challenging task for us is to generate the specific template of output with any language input or any description.

Accomplishments that we're proud of

• The Biggest accomplishment we're proud of is that we've solved a major problem for any developer.
• We have got more than 50 reviews from developers on our Instagram, LinkedIn, and Bluelearn, and they are using our site to generate readmes for their mini projects and are happy with the results.

What we learned

We have learned a lot from this single project:

• We learn how to work on a single project as a team.
• We learn about how we can use AI to ease the life of any person from any industry.
• ReadMeAI is our first project which has a dedicated link and hosting, it's also an achievement for us.

What's next for ReadmeAI

• Currently, the form of output is not under user control, we are planning to add a template marketplace where users can choose the format they want to or they can create their own template. The second most important thing is that we don't want to rely on any third-party AI LLM, so we're also working on building our own LLM specifically for ReadMeAI.
• Right now users can only generate a readme of a single file project, and we are also working on GitHub Account Integration so that user can link their GitHub account and can generate readme for bigger projects.

Conclusion

We started ReadMeAI for our side project but after getting so much positive feedback, we're seeing the potential of readmeAI as a MicroSaaS software, soon we will improve the overall UI/UX of the site and migrate to better project architecture and technologies.

Thank You

Built With

• css3
• docker
• express.js
• generative-ai
• html5
• javascript
• node.js
• showdown