Cover Letter Generator: Automate the Writing of your Cover Letters with Generative AI
Introduction
I was recently searching for a new product manager job. When applying to jobs, it’s worth it to spend some time to write a cover letter to make your application stand out. But writing cover letters is time consuming as you need to customize it for each job. It’s common to spend an hour, or sometimes more, per job application to do this well.
Fortunately, now we have Large Language Models (LLM)! And LLMs can automate or at least streamline the creation of cover letters, provided you give the right context and input.
The manual solution: writing a cover letter by hand
Before coding a software solution, it always helps to break down the manual process you’re aiming to replace.
So, how do we break down the process “writing a cover letter by hand”?
Read the job description
Identify why you’re interested - select the parts in company’s mission, culture and values that you resonate with
identify why you’d be a good fit - highlight your skills and experiences that match the job description requirements
write a cover letter to convey why you’re interested and why you’d be a good fit
The semi-automated solution: using ChatGTP prompts
Let’s first ask ChatGPT to create a cover letter with minimal context.
The result is okay, we can see ChatGPT included stuff that’s pretty common for product managers to do, but it’s of course pretty generic, as we didn’t provide much context.
Let’s include the job description for the role we’re targeting as part of the prompt.
This time ChatGPT included specific stuff that was mentioned in the job description, such as the fact that Skello recently conducted a Series B funding.
But it’s still pretty generic in the sense we’re missing some information about the job seeker: how specifically he would be a good fit in terms of skills and experience for this role. Let’s include our last professional experience and ask ChatGPT to tailor the cover letter accordingly.
Now ChatGPT is better able to connect the dots between specific requirements in the job description and our specific achievement. So the cover letter feels much more personalized now that we’ve included relevant context, compared to our first try.
This shows that the quality of ChatGPT depends for good part on prompt engineering and sufficient context.
Great! So we found a semi-automated process that works, but we want to go a bit further and not do that much copy-pasting and prompt editing for different jobs. Let’s build a more scalable solution: Cover Letter Generator.
The scalable solution: Cover Letter Generator
Why a Cover Letter Generator?
Crafting personalized cover letters for each job opportunity can be time-consuming. Cover Letter Generator automates the tedious task of cover letter writing, allowing you to generate tailored and impactful cover letters instantly.
This way you can save time, maximize impact and stay competitive.
How it works
Provide a resume (PDF file)
Provide a job description (TXT file)
Select the number of words you want to target for the cover letter
Select a tone: casual, formal, humorous…
Enjoy the result!
Of course, you can still manually tweak the cover letter that was generated. It will still be much faster than writing something from scratch every time!
How is this any different than using plain ChatGPT?
Using ChatGPT means a lot of back and forth / copy and paste, between your resume and job descriptions. A Chatbot UX is great when you’re exploring a given workflow, but less so when you want to scale that workflow. For this reason, I believe that for many workflows we will quickly move away from a chatbot UX towards streamlined LLM-enabled workflows. Writing cover letters is one example of such a workflow.
Is the result any good?
I’d say LLM really help in getting a 80% solution, and if you want to perfect it, you can tweak the remaining 20% using your human intuition to make it more convincing.
Is it somehow unethical?
Ethical concerns of using a cover letter generator can be broken down in 2 considerations: authenticity & fairness.
authenticity: using automating tools can obscure the candidate’s personality and creativity and risk devaluing human input in the hiring process. However, here I’d say that this is unfair to expect from candidates to spend several hours using their personality and creativity at the beginning of the process, while their applications are on the other hand rejected in a few seconds (or sometimes even less, with the use of AI on the recruiting side to score applications). When an application has been selected because of an initial potential fit, then it makes more sense to judge personality and creativity in the later stages of the process, without AI assistance.
fairness: it’s becoming a trope that AI will not replace humans in the short-term, but humans using AI will out-compete humans not using AI. In that sense, people using LLM for generating cover letters will develop an edge over people that don’t, in a way that will be difficult for companies to detect. This to me sounds a more solid argument, and might over time make companies abandon the expectation of a cover letter.
