Cover Letter AI

This program asks the user to put in their contact information, education and job experience, and a job description for which they want to make a application to. An API then generates a cover letter.

Comment

Inspiration

We discussed on how time consuming it took to write up cover letters for each individual company/organization we applied to. One of us thought of the idea for an AI to auto generate a cover letter. How much time would that save us?

What it does

The program asks the user to input general information found on cover letters. This includes their contact information, the businesses' name, and so fourth. The program also asks for a job description of the company which they are applying to. They simply need to copy and paste. An AI will proceed to generate a cover letter. This AI is trained based on some given job descriptions and cover letters provided before hand.

How we built it

We used the React for frontend with Express JS backend, as well as co:here API for machine learning.

Challenges we ran into

One of the challenging issues was how the API can only take a certain number of words so it made it difficult to train it under a small amount of data. The limit was 2048 word tokens. That jeopardized the quality of cover letters the API produced. Another issue we ran into was setting up the backend for React as the sample API call for co:here does not execute on React with the standard API implementation. This has cost us a significant amount of time loss trying to get the API working. We had to contact the co:here team for their documentation for React.

Accomplishments that we're proud of

We have not worked with machine learning API's for so it was a great experience both learning and working with one. We practiced our website building skills. Together it was our first Hackathon. Working as a team together was a great experience. And finally, we are proud having a completed project.

What we learned

If we organized our ideas and solutions before we actually implemented them, we would have saved a lot of time and energy. Our over all project would be more finer done and would have been better developed. Preparation is important. We both did not prepare a project idea before coming into the Hackathon so we took some time braining storming ideas.

What's next for Cover Letter AI

Using the same, perhaps another API to take in more data. Another API that could be used is to summarize the job descriptions given, and also identify the most frequent words of each job description. Getting a larger data set of cover letters and job descriptions would make the API more accurate.

Built With

Share this project:

Updates