CV Extract

CV Extractor uses LLamaIndex RAG to query your CV document with a set of query verbs such as List, Summarize and Extract. This is to assist job seekers in creating profiles on job board sites such as TradeMe and Seek.

Features

• Upload a CV document
• Query your document using the verbs and keywords.Keywords such as 'experience', 'education' and 'skills' are included
• For instance 'List Experience' will return a list of your work experiences
• Each experience is its own card whose content can be copied

Learning Goals

The lifecycle of the App is as follows:

• The user uploads CV, this document is passed to a supabase bucket.
• The user defines the query they want to carry out.
• Both the query and bucket Url is passed to the backend.
• The cv is pulled down into the server in a temp file, where it is parsed from pdf to a txt file.
• Once that is complete the server then uses LLamaIndex RAG to query the txt file.
• This extracted data is then sent up to the client.

The main goal was to learn more about RAG use and integration. Secondary goals included learning to store documents in a bucket instead of on the server and parsing pdf documents.

Demo

Here is a Demo of the App in action

screen-capture.5.webm

Set up

1. Clone the repository:

   git clone https://github.com/mano-pihema/cv-extract.git

2. Change dir:

   cd cv-extract

3. Create a .env file:

   OPEN_AI_KEY=YOUR OPENAI KEY
   VITE_STORAGE_URL='YOUR SUPABASE URL'
   VITE_ANON_STORAGE_KEY='YOUR SUPABASE KEY'

4. Install dependencies:

   npm install

5. Run App:

   npm run dev

Tech