reqspace.ai started as an AI-powered ATS for recruiters and small startups. Over time we realized the same scoring engine — pointed back at the candidate side — was a fundamentally better product than what job seekers were using. So we shipped both. Same engine. Both sides of the table.
The product started inside a recruiting practice — independent recruiters and startups stitching together LinkedIn Recruiter, a stale ATS, and a dozen tabs of Glassdoor. Scoring candidates against open reqs took hours. Most of it was rote. None of it was insightful.
We built the AI scoring engine to fix the rote work — to read a JD and a résumé and explain, in plain language, who fit and why. It turned out recruiters loved it. It also turned out the same engine, pointed at the candidate side, gave job seekers something the existing tools didn't: an honest read on how their résumé actually scored, and what to do about it.
Both products run on the same model. The recruiter side and the candidate side aren't separate companies pretending to be aligned — they're the same engine, looking at the same problem from both sides. That's what makes the candidate side actually calibrated to recruiter behavior. It's not a guess at what recruiters want. It's the same code recruiters are running.
We score candidates for recruiters and we score jobs for candidates. Same model, pointed both directions. That symmetry is the whole product — not a marketing line.
Every résumé rewrite, every match score, every "talking point" is grounded in something the candidate actually did. We don't fabricate to flatter. The trust is the moat.
When data is thin, we say so — instead of hallucinating a number that costs someone $20K or a callback. Calibration matters more than confidence theater.