Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I'm not sure if this is a good approach or not.. but I've started just exploding my resume out, then feeding it to an LLM to create a job-specific version a few times... I'll edit the job-specific version a bit, which does cut things down.

I don't know it's helped or hurt, as I've only gotten a response from about 1:50 that I sent out before or since I made the shift and I know the job market sucks.

I still need to flush out some of the prior jobs in terms of older history, projects and accomplishments. I've done a lot of contract work in 6-12 month segments in the past three decades... It's kind of wild to look back on the shear variety, scale, scope and size of some of the things I've done and worked on.

At this point, I'm not sure if it's luck, ageism or just the number of short stints in my past... but It's a weird feeling in recent job market that I haven't felt in decades. 5 years ago, it felt like I was being overwhelmed when I wasn't even looking... today it's a mess.



Last time I was job hunting I did something similar: Write out everything I've done, even "silly" things like Haskell knowledge from uni, then comment out everything not relevant to a job until my resume fits on 2 pages. My Latex Template made this much quicker than it sounds, maybe ~10 minutes per application (and 10+ hours to create the CV itself...).

Two issues I see that haven't been mentioned yet:

1. A lot of companies, especially startups, are fake job advertising. They want to look like they are growing, and they might hire a golden goose, but many job ads I saw just stay up for months or even years.

2. A lot of companies, especially large ones, are using AI to pre-screen CVs. So you now have to get through AI, then HR, then a technical manager, each with their own sets of requirements. I've played around with some of the HR AIs, they tend to be quite... superficial. To give one example from the CV above:

> Ran small A/B tests and collected human-in-the-loop safety ratings to calibrate thresholds and escalation rules.

Is a perfectly good sentence, but according to AI should be:

> Optimized escalation rules and safety thresholds by conducting A/B test collecting human-in-the-loop safety ratings, reducing false-positive escalations by 15%.

Put your achievement first? Good. Strike out verbs like 'small'? Fair, it is a sales situation. Make up numbers entirely to provide a 'quantifiable result'? Complete crap. But it seems to be what every HR bot really wants to see, so now you have to sprinkle it in and hope it gets you past the bot, and doesn't make the technical manager think you're a complete charlatan.


It's that second one that I'm trying to actually work around by using AI to generate the trimmed down version. I just haven't taken the time to pull open some of the really old versions of my resume to flush out the history and to expand on older projects yet.

The whole process just seems to suck all around. As much as I never liked filtering through a stack of hundreds of resumes as a senior member in a time hiring, being on this side of the wall is even less fun.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: