Inspiration

Many cover letter generators take minimal information and generate a terrible cover letter that

  1. Obviously is generated with AI - robotic words, inaccurate information.
  2. Provide no substantial information that hiring manager wants to see - repetitive content, too generalized. Result In 'Unfortunately...' e-mails, bad impression, frustration, time wasted for both parties.

What it does

  1. We use comprehensive information to generate a detailed, useful cover letter for the job type you are applying to.
  2. We keep history of the files you uploaded, overtime, the cover letter just gets more and more personalized for you and your desired job type.
  3. We use the Snowflake LLM technology 'mistral-large' that generates high-quality models.

How we built it

We used Streamlit to build front-end, mistral-large LLM to process user information and generate more personalized cover letter.

Challenges we ran into

  1. Streamlit user interface functionalities possibilities.
  2. Capability of snowflake-arctic LLM. (For example: max-token exceeded issue)

Accomplishments that we're proud of

  1. Allow the user the upload multiple files and extract comprehensive information from the files.
  2. Have a cool UI for user to interact with.

What we learned

  1. There is actually a lot of functionalities provided by Streamlit, Snowflake we can use in our project scope.

What's next for better-cover-letter-generator

  1. We will add more functionalities and pre-process the user information better.
  2. We will try to speed up the run time.

Built With

Share this project:

Updates