Twenty answers. No sales call required to read them.
Grouped by topic. If you have a question that isn't here, email jon@certainhr.ai — that's the direct line during the design-partner phase.
What is CertAIn, in one sentence?
AI hiring that explains itself. CertAIn reads every candidate against your JD and returns a reasoned fit assessment — ranking, interview prep, AI evaluation, and post-interview analysis — with the reasoning a recruiter can defend to a hiring manager, a candidate, or a regulator.
What are the four AI actions?
Ranking (1 credit/candidate), Interview Prep (1 credit), AI Evaluation (2 credits), Interview Analysis (3 credits). Every plan includes all four. Detail on /product.
What does "workflow-gated" mean?
Interview Analysis can only run on a candidate who already has an AI Evaluation on file. Evaluation is the baseline; Analysis compares against it. Running Analysis without Evaluation would be weaker signal at higher cost, so we don't let you.
Does CertAIn replace my ATS?
No. CertAIn is the judgment layer on top of your ATS. Greenhouse integration is the priority roadmap item; Ashby and Lever follow. Your system of record stays where it is.
Is the AI "the model itself" or a workflow around it?
A workflow around it. The model is Claude (Anthropic). Our differentiation is the structured prompts, the workflow gating, the tenant-scoped context (company profile + JD overrides + intake notes), and the reasoning-paragraph output format. Swapping models is an internal implementation detail.
How does the free trial work?
Thirty days, 100 credits, no credit card. When you click the verification link, 100 credits drop in. Pre-loaded sample JD and five sample candidates mean your first evaluation runs in under two minutes.
What happens at day 31 if I haven't picked a plan?
AI actions pause. Your account stays — you can still log in, export data, keep your JDs. Pick a plan whenever; the paid credit allocation starts fresh at the plan's monthly total (we don't stack unused trial credits into your paid plan).
What do the plans cost?
Starter $99/300 credits, Growth $249/1,000 credits, Pro $499/2,500 credits. All monthly. Full table on /pricing.
Do you offer an annual discount?
Not today. We're honest about this — we don't have enough usage data to offer a defensible annual price yet, and we don't want to lock anyone in before they've run us for a month.
What if I blow through my credits mid-month?
AI actions pause — no auto-top-up, no overage surprise. Upgrade a tier (cheaper per credit) or wait for the monthly reset.
Do you integrate with Greenhouse?
Greenhouse is the priority first integration — import candidates and JDs, push AI notes back. It's on the roadmap, not shipping on day one. The product works standalone in the meantime.
What about Ashby, Lever, Bullhorn, Loxo?
Ashby and Lever follow Greenhouse. Bullhorn and Loxo are agency-focused and are prioritized based on design-partner demand — if that's you, tell us.
Can I use CertAIn with an AI notetaker like Metaview, Otter, or Fireflies?
Yes. Paste the transcript into Interview Analysis. No integration required for the basics; deeper integrations on the roadmap.
Is CertAIn compliant with NYC Local Law 144 (AEDT)?
AEDT regulates the employer, not the vendor — so the accurate answer is that CertAIn is built to support your AEDT compliance. CertAIn does not make automated hiring decisions; every output is a recommendation to a human. We ship a bias-audit data export ready for an independent auditor, plus a candidate AI disclosure template. More on /security.
What about Illinois AIVIA and the EU AI Act?
Supported. Illinois AIVIA via the candidate disclosure template and explanation-on-request (every AI action produces a reasoning paragraph). EU AI Act (Annex III high-risk hiring) via architectural human oversight and append-only activity logging. We'll share architecture documentation under NDA for formal conformity review.
Are you SOC 2 certified?
Type II is on the roadmap, not done. Architecture is audit-ready — scoped controls, logging, access review, secret rotation, backup/DR. We'll share current control documentation under NDA for enterprise security reviews.
How is my data isolated from other tenants?
Row-level security at the database layer, tenant-prefixed file storage in R2, tenant-scoped session dependencies in the application. Super-admin impersonation is the only bypass path and is logged to an append-only audit trail. Detail on /security.
Does Anthropic train on my data?
No. Anthropic's commercial API terms exclude API inputs from their default model-training pipelines, and CertAIn does not send your data to any third-party training pipeline.
How long do you keep candidate data?
Default 24 months from last activity, configurable per tenant within what applicable law allows. Retention sweep runs daily. Deletion is soft for 30 days, then hard — a deliberate anti-resurrection window for audit.
What happens if I cancel?
Your account stays accessible for 30 days to export data. After that, hard deletion on the 30-day sweep. We send a reminder before the sweep runs. No hostage tactics, no retention call.
What about OpenAI's hiring platform?
It's announced for mid-2026 and will be an impressive distribution story. It will also not integrate with your Greenhouse, not carry reasoning between ranking and post-interview analysis, and not be owned by a team that will be on a Slack channel with you next Monday. Different product category — if you want the reasoning-paragraph workflow that explains itself, that's us.