I’ve been hiring for teams for a few years now, and I’ve heard people lament these things like you are. In practice I’ve not seen any of these “smart” scanning techniques used, it’s a recruiter comparing resumes to a checklist I gave to them (5-ish years experience, maybe a degree role dependent, or something that you think is super relevant, one of c#/java/kotlin, hiring for a mid level role so we expect some amount of experience at being self sufficient) and they filter the hundreds down to 10-15 that they screen and pass 4-5 on to me.
We did some spot checks on resumes that were passed on to make sure we were filtering ok and the quality was awful - a significant amount of people were applying for jobs asking for 5 years experience in a Java-like language with no experience, no degree and a half assed cover letter about being a good learner. A decent number were data scientists who had 2 years of python experience, and a surprising number were wildly over-qualified people who I realised after speaking to one or two they were actually trying to sell us their consulting services. That’s before you even get to “are they lying?”
Sure - that's how things are meant to work, with recruiters providing a valuable filtering service, but it does seem that many companies are now using (poor quality) AI screening, as well as a slew of abusive application practices (the ghosting. etc), and it seems any mutual respect between hiring company and candidate is disappearing. I don't know what the solution is.
My point is that that had been my experience hiring in a 10 person company and a 30k person company, and that the “suggested” AI screening isn’t happening - it’s probably that your application is the same as the other 300 applications that went in.
I haven't experienced it myself (haven't applied for a job in a long time), but there are lots of reports here on HN of people getting online applications rejected withing minutes, or late at night - definitely some companies are using software to filter resumes. This has been going on for a long time - it's not just a recent "AI" thing - resumes used to get rejected for not having the right keywords on them.
> lots of reports of people getting online applications rejected within minutes
Possibly the single most distracting alert on my phone and pc is my work email. It’s probably prioritised wrong, but if I have 200 candidates for a position, and I get an application that doesn’t meet the tech stack or YOE requirements when I have 20 who do, I’m just going to reject them.
> or late at night
ATS let you schedule emails. I used to send mine at 4am EST despite being in the UK
> definitely some companies are using software to filter resumes
I don’t doubt it, but I doubt that it’s rampant to the degree you’d believe on this site. I’d instead that it’s far more likely that the hiring manager, or a recruiter, is spending about 15 seconds looking at “does the tech stack match, how much experience, and how many other candidates are there that I think have an edge”. The people on this site are a small minority of very smart folk, but if you spend any time in a comment section of a topic you are an expert in, you’ll quickly realise that you shouldn’t take everything you read on here as absolute.
Another suggestion - Reach out to two different recruiters and get them to review your resume. (You might need to pay them to do it). You’ll get two totally different responses. Both might work, and neither might work. At the end of the day, a human makes the call, and even if the ATS is automated, a human set those criteria. Honestly, having spent so much of my time hiring over the past 5 years, it wouldn’t surprise me if there was literally no ATS scanning, and everything that was sold to fix that problem was snake oil.
What's the time frame here, what's your sample size for applying to jobs yourself, and what's your sample size for doing hiring (ie how many different companies)? I'm guessing that your personal sample size is extremely limited - not just in quantity but also in geographic area.
Of course my guess could be way off, but what you are saying is definitely the exception to the common narrative nor does it match what I've seen.
> AI screening
There's been aggressive keyword filtering since long before LLMs exploded.
I’ve been hiring in anger since 2018, applied for 1 job (sent about 10 applications) and hired over three companies. First was an 800 person company that grew to 2000 while I was there, second I was hire number 2 for a startup and I hired the full 15 person engineering team, and a decent chunk of the non engineering product/test roles into. 50 person company . Third (current role) I’ve been involved in hiring about 8 people in the last 6 months that I’ve been here. It’s a 45 person company, but a subsidiary of a 30k multinational whose hiring practices we use.
> but what you are saying is definitely the exception to the common narrative nor does it match what I've seen.
As an anecdote, I posted on who’s hiring here, and we used a separate job requisition for HN, (this was my last job where we went from 2 -> 15 people). We got about 30 applications in the 7 days following that on that req, and of those 30, only one came even remotely close to meeting the requirements on the JD - we were looking for someone with a few years experience in a Java like language, in a Europe/US time zone. Most of the candidates failed both of those criteria, hard. My point being that people who are frustrated with their situation are likely to be more vocal than someone who isn’t.
I’ve spent enough time on HN reading about topics I know a lot about, and seeing people confidently claim how X is easy or if they just Y, and they’re totally wrong. I know a decent amount about working on the hiring side - it’s been a core component of my job for the last 6 years. I’ve worked with recruiters both internal and external, spent far more time with greenhouse than any engineer should ever have to do.
My feeling is that there’s far less sophistication going on, and the dearth of human responses (which is problematic) lets people make up their own reasons as to why it’s not working when the reality is that there’s just a hell of a lot of applicants for every single job.
We did some spot checks on resumes that were passed on to make sure we were filtering ok and the quality was awful - a significant amount of people were applying for jobs asking for 5 years experience in a Java-like language with no experience, no degree and a half assed cover letter about being a good learner. A decent number were data scientists who had 2 years of python experience, and a surprising number were wildly over-qualified people who I realised after speaking to one or two they were actually trying to sell us their consulting services. That’s before you even get to “are they lying?”