ChatGPT writes. It does not hire.

Most recruiters I meet already use ChatGPT daily, and they should. It is genuinely good at the writing half of the job. The gap is everything that needs data it does not have and follow through nobody asked it for.

Start for free Book a demo

$100 of credits on us. No card required.

The short answer

Recruiters did not have a writing problem. They had a finding problem and a chasing problem, and generative text fixed neither.

Finding peopleMy own index of the open web, every name sourced

WritingVery good, with the candidate's evidence already in context

Memory of your barCalibrated on the people you moved forward and passed on

The difference in one table

	ChatGPT	Laidback
Finding people	No candidate index. Names are generated, not found	My own index of the open web, every name sourced
Writing	Very good, with a prompt each time	Very good, with the candidate's evidence already in context
Memory of your bar	Resets, or remembers loosely	Calibrated on the people you moved forward and passed on
Follow through	You copy, paste and chase	I chase, schedule and update the ATS after you approve
Where it lives	A separate tab	The channel your hiring team already reads

What to keep in the chat window

Keep it for first drafts

Job descriptions, rejection notes, interview questions, tidying a scorecard. A chat window is fast and good enough for all of it. Some of that is free on my tools pages too.

Do not trust it for names

A language model with no retrieval will produce candidates that sound right and do not exist. Anything it gives you needs verifying by hand, which is the expensive part.

It has no state

It does not know a candidate replied on Thursday, that the panel never confirmed, or that you already passed on this profile in March.

Nothing happens unless you do it

Every message is still copied, pasted and sent by you. The work I remove is exactly the work a chat window leaves behind.

#hiring-backend

Dave

I already use ChatGPT for outreach. What do you add?

Laidback

Keep using it for drafting. Ask it for twenty backend engineers in Berlin who have shipped payments infra and check whether any of them exist.

Dave

Fair. Half were made up last time.

Laidback

Because it has no index. I do. Every name I send links to the repository, the post or the paper that got them on the list.

Names with sources, not plausible names

If you already recruit with ChatGPT

Can ChatGPT find candidates with browsing turned on?

It can open pages and read them one at a time. That is not the same as an index. Coverage is shallow, results vary between runs, and there is no way to check what it did not look at.

Should we stop using it?

No. Keep it for drafting and summarising. If you want those specific jobs done for free without a prompt, the tools pages cover job descriptions, interview questions and Boolean strings.

Is Laidback just ChatGPT with a recruiting prompt?

No. The difference is data and determinism: my own crawl and profile index, plus scoring that returns the same result for the same brief. Both are things a prompt cannot add.

So use the chat window for the paragraph. Use me for the index, the scoring, the follow up and the record keeping around it.

On the independent people search benchmark I score 97.80 out of 100. The method and every scored query are published on the benchmark page.

One job, learned properly

I do recruiting and nothing else. The bar, the etiquette, the order of the stages and the follow up nobody remembers are all built in rather than configured.

I read potential, not keywords

Titles lie and stacks change. I look at what someone shipped, who they shipped it with, and how fast they grew into the work.

Every call comes with its evidence

You get the commit, the talk, the customer they carried, the line in the release notes. If I cannot show why, I do not claim it.

The same rubric, every candidate

Your bar becomes a written rubric I apply the same way on Monday and on Friday, so a shortlist is comparable instead of moody.

You can argue with me

Disagree with a score and I show the reasoning behind it, then carry your correction into the next batch.