Inspiration
In most cases, specially for small scale software projects, it's easier to build the UI first and the documentation is the boring part. So, the idea is what if we can let AI handle the documentation and we focus on what's important.
What it does
It is a command line tool which takes image as an input and generate the SRS for the given image. The image should strictly be a mobile application UI.
How we built it
The tool is built by simply using Gemini API which can take image and text as input and generate text as output. We let Gemini API do the magic and use python for I/O.
Challenges we ran into
The problem is getting the model to output in a proper format. Prompt Engineering comes to our rescue for this case.
Next problem was the character limit when generating the output. To solve this problem we limited the output scope of the document to only output requirements and a mild description.
Accomplishments that we're proud of
The tool is able to generate SRS document in Markdown format.
What we learned
We learned how to prompt Gemini API to get what we want.
What's next for SRS generation
It would be really cool if we can input multiple images to our tool to generate the SRS document and it would be even cooler if we don't have the character limitation when generating the output.
Currently, the output generated is in markdown format, it would be great if we can have the output in a docs format as used in many offices as a standard.
Next we can take requirements generated and write test cases for that set of requirements in a specified language.
Log in or sign up for Devpost to join the conversation.