There’s a new arms race running through every applicant tracking system in the market, and neither side is winning. Candidates are feeding job descriptions into Claude or ChatGPT and asking for a resume “tailored to this role” before they hit apply. Recruiters are opening inboxes full of what Bill Mastin, CEO of Cadient, calls “unicorns”: resumes that read as a perfect match because, in a real sense, they were engineered to. Mastin doesn’t mince words about where this goes: “It’s almost like Godzilla versus King Kong kind of out there battling each other,” he told me on this episode of Work Tech. One of his customers has already caught candidates using deepfakes in video interviews. Their response: going back to in-person interviews only.
That’s the backdrop for a conversation that ended up being less about one company’s product roadmap and more about what happens to enterprise software when the tuning cost drops to nearly zero. Mastin has the résumé to speak to it: a dozen years at Saba, time at PeopleFluent and Learning Technologies Group, and a five-and-a-half-year run at Topia (global mobility and compliance tech, sold last March), before taking the CEO seat at Cadient late last year. Cadient itself has a long history in recruiting technology, with high-volume ATS customers in retail and healthcare, including multi-year accounts like PetSmart and Genesco. What’s changed isn’t that history; it’s what Cadient is building on top of it.
The AI Hiring Crisis, By the Numbers
Mastin frames the current moment with a stat he attributes to Forbes and LinkedIn: something in the neighborhood of half of posted job listings may be “ghost jobs,” postings with no real intent to hire behind them. Whatever the precise number, the effect on candidates is the same, and it’s compounding a second, newer problem: volume. Six to nine months ago, mass-apply tools let candidates blast out hundreds of applications a day. Now, Mastin argues, the market has moved past a volume problem into a quality problem, because every one of those hundreds of resumes reads like a top-tier fit now that generative AI is very good at making them look that way.
He also dropped a labor-market data point worth sitting with: the share of postings requiring no prior experience has fallen from roughly one in fifteen jobs, historically, to somewhere near one in fifty today. That’s Cadient’s own hiring data, not a third-party study, so treat it as directional. Still, it tracks with what most TA leaders are already sensing about the entry-level funnel drying up.
A Modular “Smart Suite,” Built on Signal Rather Than Roadmap
The strategic core of the conversation is how Cadient is restructuring itself around this problem. The company’s ATS, what Mastin calls the “recruitment pipeline operating system,” remains the anchor for its high-volume customer base. But layered on top is what he’s positioning as a smart suite: roughly ten live modules today, spanning three categories (recruiting tech, hiring tech, and retention tech) that customers can adopt independently and plug into whatever ATS or HCM they already run, Cadient’s or not.
What I found more interesting than the module count is how deliberately unfinished Mastin says the portfolio is. He’s explicit that product-market fit varies wildly module to module: some have twenty or thirty customers, some have one, some have zero. Rather than committing a full roadmap to every idea, Cadient is using early signal, inbound interest, pilot traction, to decide where to “throw everything behind” an idea. That’s a different operating model than the traditional enterprise software cadence of annual roadmap commitments, and it’s only viable because AI has compressed the cost of standing up a new module. It also means Cadient is running multiple ICPs at once: the core ATS still sells to high-volume, vertical-specific buyers, while several of the new modules are shipping as free-trial, self-serve products, a go-to-market motion Mastin says Cadient has never used before, and one that opens the door to a smaller-volume buyer persona the company historically couldn’t reach.
Smartshield: Giving Recruiters x-ray vision on resumes
The first tool Mastin walked through live is SmartShield, a free Chrome extension available now at cadient.ai. It cross-references an incoming resume against a candidate’s public profile, employers, credentials, and other reference points, and flags where the resume and the public record diverge, including content it assesses as likely AI-generated. In the demo, it flagged four of eight checked items on a sample resume, one of them fully AI-generated.
The design choice worth noting is what SmartShield doesn’t do: it doesn’t reject a candidate or make the call for the recruiter. It surfaces flags and lets a human decide. That’s not incidental. Mastin pointed to Cadient’s published AI bias report and tied the design directly to the current wave of hiring-discrimination litigation, much of which centers on employers who let a tool make an automated decision, or skipped candidate consent, rather than keeping a human in the loop. “We put your hands on the wheel and give you automatic steering,” is how he put it. “You’re still driving the car.”
HiringScorecard.ai: An Audit Agent for the Whole Funnel
The second tool, announced the same week as our conversation, is a bigger swing: HiringScorecard.ai runs an agent-driven audit of an organization’s entire recruiting and hiring process. Feed it a company’s URL (and optionally a career page), and it evaluates the ATS in use, application completion time, screening-question sophistication, job-board distribution against industry norms, and more, then generates a scored report and a downloadable deck.
Mastin was candid that the benchmarking has real limits. Comparing a retailer’s evergreen frontline postings to a single corporate req is comparing different animals, and he described the exercise as “statistics and stories”: the numbers tell you where to look, not necessarily why. That caveat matters, because the tool’s real function isn’t diagnostic precision, it’s a lead-generation and credibility play. Cadient has built logic that takes an audit’s findings and routes the customer to a tailored solution page recommending a specific paid module (Smart Source was the example shown) as the first fix. It’s free, and it’s also a funnel.
Why Give Away the Crown Jewels?
Both tools are notable less for the specific automation than for what they signal about how Cadient, and I’d argue the broader HR tech market, is starting to compete. The free tier isn’t a stripped-down version of the paid product; it’s a full demonstration of the expertise embedded in the paid product, offered with no login wall and no sales call required. That’s a meaningfully different pitch than the traditional demo-gated SaaS motion, and it’s one that only works because the underlying agent work, resume verification, funnel auditing, used to require a consulting engagement or custom development that most TA teams couldn’t justify. Now it’s a Chrome extension and a URL field. The strategic bet is that giving away the analysis builds enough trust and habit that the remediation sale follows naturally, and that a TA leader who might otherwise cobble together their own Claude prompt or internal build will choose a vendor-backed tool with Cadient’s compliance thinking already baked in.
The Bigger Bet: Unbundling Saas
The most candid moment in the conversation, and the one I’d encourage other vendors to sit with, is Mastin’s admission about the multi-tenant SaaS era he spent most of his career building: “We’ve done a bit of a disservice to customers,” he said, describing how “customization” became a dirty word inside product organizations optimizing for one codebase serving every account. Enhancements got built, shipped, and sometimes never adopted, because they were designed for an average customer that didn’t actually exist. That’s the honest version of the 80/20 tradeoff every SaaS vendor made in the cloud-migration era: multi-tenancy bought scale and margin at the cost of fit.
His bet is that AI removes that tradeoff. Instead of one rigid platform, Cadient’s modules can be “tuned” per customer at a speed and cost that wasn’t economical twelve months ago, and, crucially, those modules don’t require displacing whatever system of record already sits underneath. Mastin’s own anecdote from his LMS days, walking into an enterprise account running seventeen separate learning platforms, is a useful reminder that large organizations have always run messier, more Darwinian tech stacks than vendors like to admit. Cadient’s modular play is a bet that the next generation of HR tech looks less like a platform war and more like a marketplace of composable agents that plug into whatever HCM or ATS backbone a company already has.
George LaRocque is the founder of WorkTech, a market intelligence and strategic advisory firm covering the HR and work technology ecosystem. WorkTech tracks global investment, M&A, and tech strategy across 60+ specialized categories. The Q1 2026 Global Work Tech VC Update and Q1 2026 Global Work Tech M&A Update are available at 1worktech.com.
Related Reading from WorkTech: