Meet the Agents That Do the Hiring Work For You
IssuePay's AI agents handle sourcing, screening, and work history verification so your team can focus on the decision, not the process.
The part of hiring nobody talks about
Most hiring tools help you track candidates. Very few help you find them, evaluate them, and verify them, automatically, without a recruiter doing the legwork.
The average engineering role gets reviewed by a hiring manager who spends 6 seconds per CV. Not because they don't care, but because there are 200 of them and 40 other things happening that week.
IssuePay's agents are built to fix that. They don't replace the human decision. They remove everything that was stopping the human from making a good one.
The three agents inside your workspace
1. Screener
The Screener reads every pending application for a job offer and scores each candidate against your screening playbook, a set of rules your team defines once.
It looks at four things, hard skills, experience, soft skills, and education, and weighs each one the way you tell it to. Hard skills means technical depth against what the role actually requires. Experience means years, seniority, and whether the candidate's scope matches the level you're hiring for. Soft skills means leadership, communication, and collaboration signals it can pick up from the CV. Education means relevant degrees, certifications, and formal training.
Each candidate comes back with a full picture, not just a number.
What this means in practice: you open your inbox and instead of 80 unread CVs, you have a ranked list, a score out of 100, and a short written summary explaining the recommendation. Underneath that, a list of strongest points with the exact evidence behind each one, and a list of weakest points, each tagged by how serious it is and paired with what the candidate would need to show to close the gap. Every required skill from the posting gets its own line, found or not found, with an estimated proficiency level. Anything worth a second look, an unexplained gap, frequent job changes, a mismatch between claimed skills and described work, shows up as a flag. You also get a one-line recommended next step, so you know whether to move fast or wait.
The Screener runs in the background. You trigger it once per job offer, and it processes your entire pending queue.
2. Sourcer
Most great candidates aren't applying to your job. The Sourcer goes and finds them, from IssuePay's own growing pool of candidate profiles, not a generic web search.
You set a candidate limit, how many to surface, and the Sourcer ranks everyone in the pool against your role. Skills match carries the most weight, followed by experience level, then location, then language fit. Candidates who've marked themselves open to work get a small boost, and so do candidates with a LinkedIn profile on file, since that's an extra layer of verification for later. Anyone below a baseline match quality doesn't make the cut at all, so you're not scrolling through people who were never a real fit.
What this means in practice: instead of waiting for applications to come in, you start with a ranked pool of pre-identified candidates, each one with a match score and a written reason attached, the same way a regular applicant would look in your pipeline. From a sourced candidate's profile, you can add them straight into the job's pipeline, or send them an outreach email without ever leaving IssuePay, pre-filled with the specific role and your company name. You'll see whether they were contacted and whether they replied, right there on their profile.
It runs in the background. You come back to a pipeline that's already populated.
3. Work History Verifier
A CV is a document someone wrote about themselves. The Work History Verifier checks it against what they actually did.
It looks at the CV on its own terms first, do the dates add up, is there a gap of several months nobody explained, does the jump from one title to the next make sense, or does it look like three levels of seniority happened in eighteen months with nothing in between. Then it checks the CV against the cover letter, looking for the two telling a different story. Then, where a GitHub or LinkedIn profile is available, it checks those too, comparing companies, titles, and dates against what's on the CV, and for technical roles, whether the actual coding activity on GitHub supports the experience being claimed.
What this means in practice: you don't get a simple flagged-or-clean stamp. You get a trust score out of 100, an overall verdict of verified, unverified, or inconclusive, and a few sentences in plain English explaining why. Below that, a claim-by-claim breakdown, specific statements from the CV, each one marked verified, likely, unverified, or conflicting, with the evidence behind it. You get a separate list of red flags and a separate list of things that actually check out, plus a cautious rating for how likely the profile is to be synthetic rather than real, because that's a real thing recruiters are dealing with now.
A concrete example of what it catches, a candidate whose LinkedIn history runs 2015 to 2019 but whose CV claims experience starting in 2021, timelines that simply don't overlap. Or a developer role where the CV describes years of hands-on coding but the linked GitHub account shows almost no public activity. Neither one proves someone is lying. Both are exactly the kind of thing worth asking about before an offer goes out, not after.
What this means in practice: by the time a candidate reaches your final round, you're not discovering surprises. You're making a decision with context.
Playbooks, your rules, their execution
Each agent runs against a playbook, a named set of rules your team defines once and reuses across roles. A playbook lets you set how much weight the Screener puts on hard skills, experience, soft skills, and education for that specific role, and two thresholds, one score to auto-advance a candidate to interview, one below which they stay in the applied pool rather than moving forward. You can turn on automatic rejection below your lower threshold, or leave every borderline case for a human to look at.
You can save more than one playbook, a senior engineering screen doesn't need to look like an entry-level operations screen, and switch between them per job.
This matters because inconsistency is one of the biggest hidden costs in hiring. Two reviewers, two different standards, two different outcomes for the same candidate. Playbooks eliminate that.
The pipeline stays yours
The agents don't make decisions. They surface information.
Every job offer starts with five boards, Applied, Screened, Interview, Offer, and Rejected, and you can add your own on top, custom named, mapped to whichever stage it should count as behind the scenes. Drag a candidate from one board to the next and it updates instantly. Each card shows the AI score where one exists, where the candidate came from, and how long ago they applied, so the board tells you something useful at a glance, not just a name.
After the Screener runs, you review candidates and move them through your pipeline the same way, just with a head start. After the Sourcer runs, the candidates it found enter the same pipeline, evaluated the same way as anyone who applied directly. After the Verifier runs, a trust card sits on the candidate's profile for as long as you need it.
Beyond scoring
A few things worth knowing that go past the three agents themselves. You can schedule an interview straight from a candidate's application, either as a native call inside IssuePay or as a Google Meet link if your calendar is connected, and the candidate gets the details by email automatically. You can message any candidate directly from their profile, sourced or applied, without switching to your inbox. And a workspace dashboard shows you, at a glance, how many applications came in, how many candidates got screened, how many got sourced, and what your agents have been doing across every open role, so you're not piecing that together yourself.
Time-to-hire, from 44 days to 20
The platform average is moving. Teams using agents consistently are closing roles faster, not because they're cutting corners, but because they're spending time where it matters, talking to the right people.
The sourcing, filtering, and verification that used to take a recruiter two weeks now happens overnight.
Getting started
If you're on the IssuePay workspace, agents are available directly from your sidebar. You'll need a job offer open, and optionally a screening playbook configured for the role.
If you're not on IssuePay yet, the workspace is built for teams that want to hire without building a full recruiting operation. Get started free or book a demo if you want to see it running on a real role.