4 ms·

Mostly been checking LinkedIn, ask HN: who’s hiring, and workatastartup It is still a pain to read through everything to try and find a match So I built a scr

by nico 3y ago

Mostly been checking LinkedIn, ask HN: who’s hiring, and workatastartup It is still a pain to read through everything to try and find a match So I built a scraper and added a bit of AI with GPT, to match my resume and ideal job description to the job descriptions It saves me hours, going from hundreds or thousands of potential matches, to only the best almost instantly The results include a summary of the role, why it’s a good fit, and instructions on how to apply Next I want to make it easier to personalize my resume for the application and keep track of the application processes (which I’m doing in a spreadsheet)

g4zj 3y ago
That's really impressive. I've been meaning to spend more time with this technology and implement it in a sample project to improve my understanding. Would you tell potential employers that you found them this way? I feel like it'd be worth mentioning if the position in question was IT-related.
- nico 3y ago
  Thank you! The interviewers don’t seem to care very much. They mostly want me to talk about big company projects with similar difficulties to what they are going through now or experienced in the past I think if I was applying to more junior positions it would be a great way to show that I can actually build something. But for more senior roles, I have the impression they want candidates to just have experience in bigger/more complicated projects that match what the hiring company is doing more closely In any case, it’s been a lot of fun building this, learning a lot and enjoying the process Here’s a short video of the interface: https://news.ycombinator.com/item?id=39562221 https://news.ycombinator.com/item?id=39562221 In the first comments there’s also links to the sources for the scraping and the gpt-processing scripts in python
  - g4zj 3y ago
    Right on. That makes perfect sense. In my experience, you're right about interviewers in that they are far more concerned with matters relevant to the position, as well as their company and its projects. And rightfully so, unless they are concerned that the candidate may be too inexperienced to actually build something from the ground up as you have. The project looks very interesting. Nicely done! I'd really like to learn to put the pieces together for such a project. Too focused on other things at the moment however.
    - nico 3y ago
      Thank you for checking it out! ChatGPT helped a lot. I had no experience working with ncurses for example, and gpt has been great at providing boilerplate code, as well as at iterating over smaller bits of code, to refactor or add features It’s amazing how fast you can go if you can rely on some intuition from your experience but use the speed of the LLMs to iterate quickly with coding, especially when using something new

by nico 3y ago

g4zj 3y ago

That's really impressive. I've been meaning to spend more time with this technology and implement it in a sample project to improve my understanding. Would you tell potential employers that you found them this way? I feel like it'd be worth mentioning if the position in question was IT-related.

nico 3y ago
Thank you! The interviewers don’t seem to care very much. They mostly want me to talk about big company projects with similar difficulties to what they are going through now or experienced in the past I think if I was applying to more junior positions it would be a great way to show that I can actually build something. But for more senior roles, I have the impression they want candidates to just have experience in bigger/more complicated projects that match what the hiring company is doing more closely In any case, it’s been a lot of fun building this, learning a lot and enjoying the process Here’s a short video of the interface: https://news.ycombinator.com/item?id=39562221 https://news.ycombinator.com/item?id=39562221 In the first comments there’s also links to the sources for the scraping and the gpt-processing scripts in python
- g4zj 3y ago
  Right on. That makes perfect sense. In my experience, you're right about interviewers in that they are far more concerned with matters relevant to the position, as well as their company and its projects. And rightfully so, unless they are concerned that the candidate may be too inexperienced to actually build something from the ground up as you have. The project looks very interesting. Nicely done! I'd really like to learn to put the pieces together for such a project. Too focused on other things at the moment however.
  - nico 3y ago
    Thank you for checking it out! ChatGPT helped a lot. I had no experience working with ncurses for example, and gpt has been great at providing boilerplate code, as well as at iterating over smaller bits of code, to refactor or add features It’s amazing how fast you can go if you can rely on some intuition from your experience but use the speed of the LLMs to iterate quickly with coding, especially when using something new