Twelve Stack
← All case studies
Intelligent triage

Every inbound application, pre-scored before a human reads it.

Industry: Government contracting Build time: ~2 weeks Volume: 8,000+ candidates screened to date

A firm hiring across dozens of government contracting roles at once was drowning in resumes. Today, every application is scored, summarized, and ranked before a recruiter opens it — and the narrative tells them what to ask before the phone screen.

The outcome

  • Recruiters open their queue to a pre-sorted list — the strongest candidates surface immediately.
  • Every application gets a 0–10 fit score plus a short narrative on strengths, gaps, and what to probe in a screen.
  • Reference quality is scored automatically, separating professional-grade references from personal ones.
  • The team spends less time reading resumes and more time on the candidates that matter.
  • The system runs in the background — nobody has to remember to use it.

The situation

The firm hires across dozens of government contracting roles at once. Every job has its own technical criteria, clearance requirements, and must-haves buried in the description. Applications were coming in faster than anyone could read them, and there was no way to tell a 3 from a 9 before opening the file.

They had just migrated from ADP Recruitment to Pinpoint as their ATS. That transition was the opening to build something on top of the system rather than around it.

What the system does

When an application lands in Pinpoint, an AI workflow runs automatically. It reads the job posting, extracts the actual hiring criteria (not just keywords), scores the candidate against those criteria, and writes the result back into a custom field recruiters already see in their normal review queue. A short narrative summarizes strengths, weaknesses, and the gaps worth asking about.

Pinpoint candidate view showing AI-generated strengths, weaknesses, gaps, and a fit score written back into custom fields.
AI-generated strengths, weaknesses, gaps, and fit score written back into Pinpoint custom fields.

A second workflow evaluates the references each candidate provides. Professional addresses and verified business relationships score higher than personal emails or family contacts — so recruiters see at a glance which references are worth a call.

AI Reference Score and accompanying notes assessing the credibility of a candidate's references.
The reference-quality workflow flags weak references — personal email domains, vague naming, missing tenure — alongside the stronger ones.
"The pattern here — AI doing the first pass on high-volume inbound, writing structured output back into the system the team already lives in — works anywhere there's more incoming work than people to read it. It's one of the highest-leverage things a small team can build."

Have a triage problem like this?

Start with a 15-minute conversation. No pitch, no pressure — we just figure out whether there's something worth building.

Book a call