The HireQwik Dashboard: What HR Sees After an AI Screen

An AI screening dashboard for HR review is supposed to save a recruiter time, but most vendor dashboards show you exactly what a spreadsheet already shows: a name, a score, a status column. We built the HireQwik dashboard to answer a narrower question: out of everyone who interviewed since you last logged in, who actually needs your attention? Here’s what loads when a TA lead opens /dashboard/hr the Monday after a weekend campus drive.

The four bands, and what decides which one a candidate lands in

Every candidate who completes a HireQwik interview is sorted into one of four bands: Strong Go, Go, On Hold, or No Go. The sort isn’t a resume-match score dressed up with a new label. It comes out of a 15-20 minute structured voice conversation, scored in the order communication quality first, then role fit, then domain basics. That ordering is deliberate. A candidate who can’t hold a coherent conversation about the role is a No Go regardless of what their resume claims about the role, and a candidate who communicates well but is thin on domain basics usually lands On Hold rather than getting auto-rejected, because that’s exactly the profile a hiring manager sometimes wants to see anyway.

The dashboard doesn’t ask a recruiter to read four bands of raw output and decide what they mean. It’s already decided what each band means. Strong Go and Go are ready to move forward. No Go is a clear pass, with the transcript attached in case anyone asks why. On Hold is the only band that requires a human judgment call, and it’s built to be small.

The four bands at a glance

Why the queue is already thin by the time a human opens it

The reason On Hold stays small traces back to a 3,000-candidate pilot campaign that we ran in a single evening early in HireQwik’s rollout. Across that campaign and the ones since, 60%+ of candidates score as clear No Go: mismatches that a resume alone rarely flags cleanly, because a resume doesn’t test whether someone can actually explain their own experience out loud. That auto-sorted 60% is why the same review that used to take a recruiter roughly 18 hours per drive now takes 1-2 hours: 89% less time, not because the AI works faster than a human, but because the human is no longer starting from an unsorted pile.

Open the dashboard after a weekend drive and the arithmetic is visible in real time. A campus drive that pulled a few thousand applicants doesn’t hand HR a few thousand rows to triage. It hands them a review queue that’s already down to the fraction genuinely worth a human decision, which is the only part of the job a recruiter’s time should be spent on in the first place.

Every score is backed by a transcript, not just a number

Click into any candidate in the On Hold band and the score isn’t a black box. The exact exchange that produced it sits directly below the number: the question that got asked, the answer the candidate gave, and where the scoring rubric drew the line. That matters practically, not just for auditability: two recruiters looking at the same borderline candidate can actually disagree about the call, and having the transcript in front of both of them means the disagreement is about the candidate’s answer, not about what the AI supposedly heard.

It also means the same rubric gets applied to candidate one and candidate three thousand, at 6 AM and at 11 PM, without drift. A human phone-screener’s bar shifts across a long day whether they notice it or not. The dashboard’s bar doesn’t.

Reading a single candidate row before deciding

A band tells you where a candidate landed; the row tells you why, and the two other things worth glancing at before making a call both sit on it. The first is the resume-match score the candidate carried into the interview — a different number making a different claim, and when the two disagree, that disagreement is information, not an error: a strong resume that produced a weak conversation is precisely the case the second stage was built to surface. The second is the specific exchange behind the verdict — not the whole transcript, but the question the rubric flagged as the binding one. For an On Hold candidate that’s usually a single answer, and reading it takes less time than forming a secondhand opinion from the score alone.

The habit worth building is reading the row in that order — band, then the score it disagrees or agrees with, then the one exchange that decided it — because that’s the order in which each piece narrows what the next one means.

What the dashboard won’t do for you

The four bands sort on communication and role fit. They don’t replace a hands-on skills round for work where that’s the actual bar. A manufacturing shop-floor hire or a role that lives or dies on a technical assessment still needs that assessment; the dashboard tells you who’s worth taking to that next step; it doesn’t administer it.

The other honest limit is operational, not technical: an On Hold queue that nobody checks for a week isn’t a filter, it’s a backlog with a nicer name. The dashboard makes the queue small enough to clear in a sitting. It doesn’t clear it for you. Teams that treat On Hold as an inbox with a same-day habit get the benefit of the sort; teams that let it sit are back to reviewing a pile, just a smaller one.

Running the dashboard as a daily habit

Teams that get full value from the four bands treat the dashboard like an inbox with a same-day rule, not a report they visit when someone asks for numbers. The rhythm is short: open it once after each drive day, clear On Hold in a single sitting while it’s still small, and batch the Strong Go and Go bands forward before they age — a Strong Go that sits for a week is a candidate another company hired. The No Go band needs nothing on a normal day; its job is to be ready with a transcript on the day a hiring manager, a candidate, or an auditor asks about one specific name. Run that loop consistently and the dashboard stays what it was built to be: the shortest path between an interview finishing and a human doing something about it.

A dashboard that bolts an “AI score” column onto a spreadsheet isn’t screening — it’s decoration. The bands only earn their keep if the recruiter opening them can trust that Strong Go means something and that No Go was actually a fair call, transcript included. That’s the bar we built the dashboard to clear, not just the bar of looking automated.

If your team is still triaging a raw applicant spreadsheet instead of a sorted shortlist on a Monday morning, see what the dashboard actually looks like at your hiring volume.