Dover
Dover is an AI recruiting platform for startups offering a free applicant tracking system, candidate sourcing via a LinkedIn/Copilot browser extension, OpenAI/Anthropic-powered 0-100 applicant scoring, and a fractional-recruiter marketplace.
§ 01 — Score breakdown
§ Score breakdown
Category scoring
Weighted contribution shown to the right of each bar.
- 01
Article 11 Technical Documentation
Weight 20%34
+6.8
- 02
Bias Audit Transparency
Weight 18%52
+9.4
- 03
FRIA Support
Weight 15%30
+4.5
- 04
Data Governance Disclosure
Weight 15%50
+7.5
- 05
Human Oversight Design
Weight 12%58
+7.0
- 06
Post-Market Monitoring
Weight 12%44
+5.3
- 07
Customer Documentation
Weight 8%60
+4.8
§ 02 — Strongest · weakest
Strongest category
Customer Documentation
Raw score 60 · contributes 4.8 to total.
Weakest category
FRIA Support
Raw score 30 · contributes 4.5 to total.
§ 03 — Cited evidence
Download diligence record→§ Evidence
Cited per category
Every score is backed by at least one cited piece of evidence.
§ 04 — Editorial notes
Company overview
Dover (legal entity Staya Inc., dba Dover) is a San Francisco-based recruiting platform founded in 2019 and part of Y Combinator's Summer 2019 batch, having raised roughly $23M and served over 1,500 startups. Its stack combines a free ATS, a browser-based sourcing/outreach extension (Copilot), AI-powered applicant scoring that uses a customer-supplied OpenAI or Anthropic API key to rank resumes 0-100 against custom criteria, and an on-demand marketplace of fractional recruiters. The AI is positioned as a prioritization aid rather than an autonomous decision-maker.
Regulatory exposure
Dover's AI applicant-scoring and sourcing tools are automated employment decision tools within scope of NYC Local Law 144 and fall under Annex III high-risk employment use of the EU AI Act, plus emerging Illinois and Colorado AI-hiring rules. Dover has engaged the NYC regime directly: it published a customer-facing LL 144 explainer and posted a bias-audit summary, and it frames scoring as no-auto-reject decision support with mandatory human review, which reduces but does not remove exposure. However, the single public audit dates to June 2023 with no visible annual renewal, and there is no EU AI Act deployer material, so its documented posture lags current obligations.
Path to a higher score
The fastest wins: renew the 2023 bias audit annually with a recognized independent audit firm (rather than an individual) and post the current-year selection/impact-ratio summary, restoring LL 144 currency. Publish an AI system/model card and a responsible-AI/explainability statement to build Article 11 evidence, and add EU AI Act Art. 27 FRIA templates and deployer guidance. On data governance, obtain SOC 2 Type II or ISO 27001/42001 attestation and state explicitly that candidate data is excluded from model training. Finally, add an AI-specific model-update changelog and continuous monitoring surface beyond the uptime status page.
Conflicts of interest
No vendor pays for placement, scoring, or removal. Casework — the consulting firm that operates this directory — provides paid services to some vendors. Any active or recent (within 24 months) commercial relationship is disclosed on the affected vendor profile and the review is reassigned to an independent reviewer. See the full policy on About.
Casework has no commercial relationship with this vendor.