Skip to content

Writing emails to apply for jobs is TIRING. So i invented the solution!

License

Notifications You must be signed in to change notification settings

ahmed-debbech/email_job_ai

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Email Ai

Write your next job application with Ai. Sounds common ?
Here is why it is not:

  • You copy paste the job description if any to it.
  • Based on a template prompt you write once about who you are.
  • Email Ai will combine the template with the job description to produce a fully generated Email for you.

Also it is private because you need to host it on your own. Keep in mind Email Ai is not a public platform, it is only a self-hosted one that you will use privately (see bellow on how to easily set it up)
PS: You need to install Ollama as the ai infrastructure and you can use any model.

How frequent you're going to need it?

When you are lazy to write an email, you press a button.

How to use it ?

0- (you do this after fresh install): You visit http://your-host-ip:8087/api/config.html to configure (Ollama IP, Ollama Model, your template)

1- You get the job description. (if any, if not it is okay too).

2- You past the job description to the system through an IOS shortcut (we shared bellow)

note : for now you can use the REST API or the IOS Shortcut, soon we will develop more clients. (more about the API bellow)

3 - You visit http://your-host-ip:8087/api/index.html to see all your done/pending write requests

The IOS Shortcut will ask you to provide:

  • job link (optional)
  • job name, used to keep track of requests later
  • job description (optional)
  • things to add to the email (optional) (like: write it in frensh, say that i have already h1b visa, etc...)
  • template number to use (default is 1)

How to install

  • For server: build the jar file or you can use our docker image with docker pull the1dev/email_ai:latest.
  • For IOS Shortcut client: you find here the link to install it on your Iphone.
  • Install Ollama preferably on same machine and pull your fav model to use (llama3.2 is recommended).

Technical stuff

1- Configuration:
Set the ollama server ip:port and set the ollama model to use like llama3.2, Don't forget to add your template

Here is an example of a template prompt
Write for me an email about a job i found i am X Y with 5 years of Java experience and i have worked on tada tada and tada, you can find my github here link_github and my linked in on link_linkedin. make the email fits this job description " &&job_desc " and please also add that: " &&set_mine " make the email unique and combines both my profile and job description

note that &&job_desc and &&set_mine are necessary placeholders that must be set in the template to be able to replace it with job description and optional addition that you wrote in the request.

2- Endpoints: \

  • POST /api/write

    {
    "job_link": "https://www.linkedin.com/jobs/collections/recommended/?currentJobId=1",
    "name": "The name",
    "job_desc": "Degree in IT Engineering or similar areas At least 3 years of relevant professional experience in Java programming language Experience with Spring or Spring Boot Experience with Frontend development using Angular Experience with SQL Server Good level of English (B2-C1)",
    "set_mine": "add that i have CCNA certificate, and write in portegeuse",
    "template_num": 1
    }
    

    any client should consume this endpoint for sending write requests

  • GET /api/read
    this endpoint is used to see the pending/done requests.

Open source, so why not to contribute?

Feel free to open a pull request or contact me at my email about changes you want to see.
As this project is still under developement but acquiring a finished demo you can try.

Credits

Ahmed Debbech - Original author