Cold Mail generator

we made cold mailing a lot more easier then before.

Comment

Inspiration

For the past couple of months, we were looking for internships and we have been cold mailing a lot of founders as well as HR, we realized that cold mail cannot be drafted in one general way since different companies have different visions and we have to pitch how we will add value to their company. This is a tedious task, we came up with a hack that auto-generates cold mail and makes your life simpler.

What it does

It basically takes a set of inputs from the user(example: the designation that they are applying to, their skills and projects and experience) in the backend our platform makes a call to the GPT 3 API and gets back a full-fledged mail in seconds and navigates the user to the mail interface and allowing him to send the mail to the person in one click!

Don't worry we have backed up lazy folks like us who can use web kit audio-to-text tool to customize their cold mails.

How we built it

This is a simple web application where the front end is built using HTML, CSS, javascript and ejs templates and at the backend we use node.js, express.js which help in server routing and API calls. we have used chatGPT 3 API

Challenges we ran into

1st challenge The main challenge was parsing the json, the response we were getting from chatGPT for the body of the mail had a lot of escape sequences as well as few overriding quotation(example: I'm which made json parsing not possible, since we used ejs, we could get old of each and every object member that can be seen in code)

2nd challenge front-end challenge. we are a group of blockchain developers who are not well versed with the front-end technologies, it was a bit challenging for us to retrieve and send the form data as well as animation problems.

3rd challenge Our team member had to research a lot on how to maximize the prompt level for getting the best result from GPT.

4th challenge Audio to text challenge we had to figure out the best way to convert the audio to text so as to reduce the API call loads on the server, otherwise the application would be slow. So, we used "webkitSpeechRecognition" tool available in the windows object of javascript to make things easy.

Accomplishments that we're proud of

we had no much knowledge about AI technology but by leveraging openAI API docs we were able build such a good use case application for a hustler mind set person. we completed this project within the time constraint and made us realize we have improved a lot more in couple of months. Has a team we understood to split our works into code writing, research, designing and project management, making us realize how good we can work together.

What we learned

AI is not the next big thing, perhaps it's already a big thing! It very important for a developers to leverage AI to build applications that will make human life easier. The key learning that we take from this hackathon are 1) Diversifying the work 2) Time management 3) Team coordination 4) Mindest to overcome the challenges and go on building wonderful products

What's next for Cold Mail generator

This project is a very small piece of the bigger picture where we cannot disclose all the future plans but I'll tell you two extended features of this. 1) extending our product to write effective cover letters 2) To make a automated ecosystem where the user has to just put his information, our system will automate and send mails to related people from our database. We believe that we will use GPT4 to accomplish this automated ecosystem of sending mail

Share this project:

Updates