From 500 applications
to five worth calling.

Alcax Technologies builds workflow automation software for mid-market operations teams. Based in Austin with a remote-distributed engineering org, they were growing headcount by roughly 30% year over year when this story takes place — adding engineers, product managers, and customer success hires simultaneously.

Their talent team at the time was two people: a recruiter and a talent acquisition lead. Fourteen open roles. No additional headcount approved for HR.

The pipeline was full. The candidates weren't right.

Alcax was pulling sourcing from three places: LinkedIn Recruiter, Indeed Sponsored Jobs, and their own careers page. For a senior backend engineering role they posted in Q3, they received 470 applications in the first two weeks. Their recruiter screened 60 of them before the recruiting lead called it — the hit rate wasn't good enough to justify the time.

The core issue wasn't their ATS or their job descriptions. It was that none of the sourcing channels had any incentive to send qualified candidates — only to send volume. Someone applying from a completely different industry looks the same in an ATS inbox as a strong candidate until a human opens the file.

They tried tightening keyword filters in Greenhouse. That helped some, but they started noticing strong candidates getting filtered out because their résumés didn't use the exact phrasing the filter expected. The tool that was supposed to save time was now creating its own errors.

“I was spending three hours a day just doing first-pass triage. Not sourcing, not interviewing — just deciding who was clearly not right. That can't be what this job is.”

Matching happens before the application. Not after.

Alcax connected JobMinglr to Greenhouse using the native integration. They submitted their job board URL and API key through the self-service form — the setup took about 20 minutes, and their open roles were live in JobMinglr's candidate-facing feed the same day.

What changed structurally: JobMinglr doesn't show a candidate a job until the algorithm has already evaluated whether they're a reasonable fit. Skills, experience level, salary expectations, location, work style preference — all of that gets scored before the job surfaces in someone's deck. A candidate who swipes right has already cleared a baseline match threshold. They chose to apply knowing what the role looks like.

On Alcax's end, nothing changed in Greenhouse. Candidates still came in as standard applicants. The difference was that each one arrived with a match score and a trust score already attached — so the recruiter could see at a glance who the strongest candidates were without opening every résumé first.

They ran JobMinglr alongside their existing channels for the first six weeks — not as a replacement, but as an additional source they could benchmark against.

“By week three we could see the difference just from the intake quality. We weren't doing anything different in Greenhouse. The candidates coming through JobMinglr were just better — more specific in their interest, further along in their thinking about the role.”

Three months of data.

Alcax tracked results across the first full quarter after connecting JobMinglr, comparing against their trailing three-month baseline from traditional sourcing channels.

Unqualified applicants reaching the review stage dropped 42%. That number came from comparing first-pass rejection rates between JobMinglr-sourced candidates and candidates from LinkedIn and Indeed over the same period. Time-to-fill shortened by 35% — measured from role opening to accepted offer.

The engagement number — 2.3× — reflects a combination of higher recruiter outreach response rates and better interview show rates. Candidates who came through JobMinglr were less likely to ghost scheduling requests and more likely to show up prepared. The working theory from their talent lead: people who went through a matching process before applying were more deliberately interested, not just broadly applying.

The recruiter recovered roughly two hours per day previously spent on triage. That time went back into sourcing for harder-to-fill roles and improving their interview process. The team closed nine of fourteen open roles during the quarter — their highest quarterly close rate since the company was at 80 people.