What it is
Crosschq is a hiring-intelligence platform: it instruments the hiring funnel (résumé matching, AI interviews, assessments, live-interview capture, 360 reference checks, fraud and identity checks) and joins that pre-hire signal to post-hire outcomes, so a recruiting team can say which sources, interviewers and scores actually predicted a good hire. Most buyers meet it through reference checking (Crosschq 360). The quality-of-hire analytics layer is what it is built to sell.
It is no longer a standalone company. On 17 September 2026 Crosschq, ApplicantX (candidate fraud and identity) and Traitify (behavioral and cognitive assessments, acquired by Crosschq on 31 March 2026) were combined into Brilliant AI (brillianthire.ai), with a strategic investment of undisclosed size from Bessemer Venture Partners. Mike Fitzsimmons, Crosschq’s CEO, runs Brilliant. The three brands keep their names “in the near term”, and Brilliant positions itself as the orchestration layer above them. Brilliant reports 500+ customer organizations, 40+ HR and talent-system integrations, and models trained on 200M+ hiring outcomes; all three figures are vendor-reported.
The Crosschq product menu now runs in three groups. Evaluate: Realistic Job Previews, AI Match résumé matching, AI Agent interviews, Traitify assessments, Interview IQ for live interviews, and 360 references. Protect: candidate fraud detection (built on ApplicantX) and ID verification powered by ID.me. Measure: Insights for quality of hire and recruiting metrics, the TalentWall pipeline view, and Quin, a plain-English query assistant over hiring data.
Why it shows up in Recruiting stacks
- It closes the loop most stacks leave open. Greenhouse, Ashby and Workday record what happened in the funnel. Insights joins that to performance and retention data pulled back from the HRIS, which is the join nearly every team plans and almost none finishes.
- Fraud and identity are a product line, not a feature. ApplicantX scans applications arriving through the ATS for identity-fraud signals and writes results back to the candidate record. Deepfake and impersonation checks cover the interview, and ID.me handles document-backed identity. Crosschq’s own customer data puts detected fraud at 11.7% of applicants, a vendor figure, but the direction matches what TA teams report.
- References become structured data. 360 collects references through a questionnaire and returns scored competency data, so a reference becomes a comparable field rather than an anecdote. The split against the reference-check summary skill is clean: buy 360 for async collection at volume, run the skill if your hiring managers insist on live calls.
- There is now an agent surface. Brilliant MCP, announced with the September launch, exposes Brilliant’s screening, scheduling and verification models to enterprise AI agents over the Model Context Protocol. That matters if your recruiting agents already run in Claude or a Workday agent layer.
Pricing
Quote-only. There is no pricing page (crosschq.com/pricing returns a 404), no free tier, and every call to action is a demo request. Vendr’s buyer guide reports an average contract of $9,663/year across 5 deals, with observed contracts running $2,500 to $18,000/year. Five deals is a thin sample, so read it as a band, not a benchmark. For a single module (360 references, or fraud screening on one ATS) the band holds. Once AI interviews and Traitify assessments are in scope, pricing is volume-driven and the contract moves out of that band; ask for the per-assessment and per-interview rate card before the platform quote, because that is the number that scales.
Best for
Recruiting operations leads and TA analytics owners at 200-to-5,000-person companies, or high-volume frontline hirers, who already run a modern ATS, can get at HRIS performance data, and need to defend hiring decisions on quality of hire and fraud exposure rather than time-to-fill.
Not for teams hiring under ~50 people a year: the quality-of-hire model needs outcome volume to say anything, and a reference tool plus a manual review is cheaper.
Alternatives and when to pick them instead
- HireVue: the enterprise incumbent for assessment plus video interviewing. Pick it when procurement wants one large validated vendor with published adverse-impact studies, and you can absorb the rollout.
- SHL: the psychometric incumbent. Pick it when assessment validity is itself the deliverable and legal expects to defend it in a hearing. Traitify is credible but narrower.
- Mercor and Apriora: fast-growing AI-interview entrants. Pick them if screening throughput is the only problem and you do not want the analytics, assessment or reference surface.
- iCIMS: owns SkillSurvey. Pick it if you are already an iCIMS shop and reference checking is all you need; the native path beats a second vendor on price and on data movement.
- Checkr: pick it when the fraud question is really a background-screening and compliance question; Crosschq’s fraud layer catches impersonation earlier in the funnel but is not a consumer reporting agency.
- Eightfold: pick it when the requirement is talent intelligence and internal mobility across the employee lifecycle, not hiring-outcome measurement.
The fallback is a quarterly manual quality-of-hire review joining ATS exports to HRIS performance ratings in a spreadsheet. It is tedious, it works, and it tells you whether the automated version is worth buying.
Watch-outs
- You are now buying from a roll-up, not a point vendor. Brilliant is three products under one owner, with brand names that will “gradually” fold into one ecosystem. Contract names, SKUs, support routes and data-processing terms can all change mid-term. Guard: contract with the legal entity that will exist at renewal, require notice and a no-cost exit if the product you bought is renamed or bundled, and get the data-processing agreement to name every Brilliant entity that touches candidate data.
- The outcome model is vendor-reported and unaudited. The “200M+ hiring outcomes” figure sits beside a 170M+ figure that Crosschq’s own product copy has also used; the vendor does not reconcile the two. Guard: treat predictive scores as a ranking aid, not evidence. Before renewal, backtest predicted quality of hire against your own 12-month performance and retention data, and require that correlation in writing.
- Reference response latency is the recurring user complaint. G2 reviewers repeatedly flag references who never reply, which stalls the offer stage. Guard: never gate an offer on reference completion; set a 3-business-day SLA and a documented override path before rollout.
- Brilliant MCP is announced, not proven. It is a September 2026 launch at the parent level. Guard: confirm in writing which Crosschq modules it reaches, how it authenticates, and whether agent calls are logged against the candidate record before you design a workflow around it.
- AI interviews and assessments are regulated hiring tools. NYC Local Law 144, Illinois AIVIA and the EU AI Act all apply. Guard: obtain bias-audit artifacts for the AI interviews and for the specific Traitify instruments you deploy. A platform-level compliance statement is not an audit of the module you turned on; consult counsel on what your jurisdiction requires.