The Hidden Cost of Hiring Blind
Resumes tell you what someone did. They do not tell you who they are, and that is the expensive part.
Why Traditional Hiring Fails
The Resume Problem
- - Lists accomplishments, not behavioral patterns
- - Optimized for keywords, not compatibility
- - Tells you where they worked, not how they work
- - No insight into values, motivations, or adaptability
The Interview Problem
- - Candidates rehearse answers, not behaviors
- - Unconscious bias skews evaluations
- - 30 minutes cannot reveal personality dimensions
- - No standardized way to compare candidates
ProvenATS solves this by analyzing the data candidates already leave behind, their LinkedIn profiles, to build a complete psychographic picture before you ever talk to them.
A complete psychographic stack
Up to 22 Dimensions
Three tiers: Standard (11), Advanced (17), or Premium (22 dimensions). From core personality to full psychographic profiling with career anchors and SDT needs.
60-Second Analysis
From LinkedIn URL to comprehensive psychographic profile in under a minute. No questionnaires. No scheduling. No waiting.
Live Coding Assessments
Token-invited browser sessions with a Monaco editor, real terminal, and AI-generated problems tailored to the role. Graded across nine dimensions of process and craft.
Interview Q&A Scoring
Save interview questions and answers per candidate. AI grades responses against the role and returns a verdict, overall score, and per-question reasoning. Surfaced on the comparison view.
Role-Specific Scoring Controls
Set tier weights and per-dimension minimum targets per job. Candidate fit scores recompute on read so changing the controls updates rankings without re-running inference.
Candidate Library
Every analysis joins a persistent pool with tags and notes. Re-score the entire library against a new job description in one pass; compare any 2 to 6 candidates side by side.
One-Click Rejection
Capture the candidate's email at analysis time, encrypted at rest. Editable rejection-email template per account. One click sends with delivery audit trail.
Confidence Scoring
Every insight includes a confidence score based on available data. Know when to trust the analysis and when to dig deeper.
Cost to Solve It
One investment. Unlimited returns. Compare the cost of ProvenATS to just one bad hire, and see why the math works in your favor from day one.
Starter
- Up to 500 analyses/year
- White-label branding
- Standard support (48hr)
- Basic ATS integrations
- Standard (11 dimensions)
Professional
- Up to 2,000 analyses/year
- White-label branding
- Priority support (24hr)
- Full ATS integration suite
- Custom integrations (1)
- Batch processing
- Advanced (17 dimensions)
Enterprise
- Unlimited analyses
- Dedicated managed instance
- Dedicated support (4hr response)
- Unlimited custom integrations
- SLA guarantees
- Optional: BYO Infrastructure (+$10K)
- Premium (22 dimensions)
Need More Analyses?
All tiers include: Full source code, Managed hosting, Security updates, Feature upgrades
Enterprise: Choose managed (we host) or BYO Infrastructure (you host on your accounts)
ProvenATS vs. The Alternatives
See how we compare to traditional hiring tools and generic AI solutions.
| Feature | ProvenATS | Traditional ATS | Generic AI Tools |
|---|---|---|---|
| Data Source | LinkedIn + Behavioral | Resumes only | Varies |
| Assessment Dimensions | Up to 22 psychographic dimensions | Skills & experience only | Generic summary |
| Analysis Speed | 60 seconds | Manual review | Minutes to hours |
| Candidate Experience | Zero friction (no questionnaires) | Application forms | Often requires input |
| Scientific Foundation | Validated psych frameworks | None | Rarely validated |
| Confidence Scoring | Every assessment | N/A | Rare |
| Architecture | 14+ specialized inference agents | Keyword matching | Single LLM prompt |
| Ownership Model | One-time license, own forever | SaaS subscription | API credits / subscriptions |
| White-Label Option | Included | Expensive add-on | Rarely available |
vs. Traditional ATS
ATS systems organize candidates. We analyze them. While your ATS tracks where candidates are in the funnel, ProvenATS tells you which ones are worth moving forward, based on who they are, not just what they have done.
- Complements (does not replace) your ATS
- Integrates with Greenhouse, Lever, Workday, and more
vs. Generic AI Tools
Most AI tools use a single prompt to analyze candidates. We use 14+ specialized agents, each calibrated on validated psychological frameworks (HEXACO5Ashton & Lee (2007), Personality and Social Psychology Review: foundational paper establishing HEXACO as a six-factor personality model with empirical and theoretical advantages over the Big Five.journals.sagepub.com, Schwartz6Schwartz (1992), Advances in Experimental Social Psychology: foundational cross-cultural framework establishing 10 universal human values across 20 countries.psycnet.apa.org, RIASEC7Holland (1997): foundational theory of vocational interests and work environments, the basis of the RIASEC model.psycnet.apa.org, Schein, SDT). The result is evidence-cited scoring against published instruments rather than a single freeform summary.
Independent meta-analytic research finds the HEXACO model explains roughly 60% more variance in counterproductive workplace behavior than the legacy Big Five (31.97% vs 19.05%)9Pletzer et al. (2019), Journal of Vocational Behavior meta-analysis: HEXACO explains 31.97% of variance in workplace deviance versus 19.05% for the Big Five — roughly 60% more predictive power.doi.org/10.1016/j.jvb.2019.04.004. Structured behavioral assessment (r=.42) also predicts job performance more than twice as strongly as unstructured interviews (r=.19) in the most current meta-analytic synthesis of selection methods8Sackett et al. (2022), Journal of Applied Psychology meta-analytic update: structured behavioral assessment (r=.42) predicts job performance more than twice as strongly as unstructured interviews (r=.19).doi.org/10.1037/apl0000994.
- Domain-specific training data for each dimension
- Cross-validated results with confidence scoring
How It Works
Multi-agent architecture. One URL. Complete profile.
The Process
- 1.Submit a LinkedIn URL plus optional job description, candidate email, and resume text through the dashboard or API
- 2.Our system fetches public profile data, posts, and activity, and parses the job description into structured fields
- 3.A panel of 14+ specialized agents analyzes psychological dimensions across 4 tiers, gated by Tier 1 structural transition detection
- 4.Results cross-validate, merge into a unified profile, and a Sonnet-generated summary card returns a verdict (strong fit / review / not recommended)
- 5.Capture interview Q&A on the candidate page; the AI scorer returns a verdict and per-question reasoning
- 6.Compare 2 to 6 candidates side by side; tune tier weights or set dimension targets per role and rankings recompute on read
Data Sources
- LinkedIn profile, headline, and summary
- Full work history across every employer
- Posts, comments, and reactions (where public)
- Skills and endorsements
- Education and certifications
- Public GitHub portfolio (optional enrichment)
- Optional candidate resume text (encrypted at rest)
- Captured interview Q&A pairs (post-analysis)
Security & Compliance
Frequently Asked Questions
How is this different from a traditional ATS?
Traditional ATSes organize and track candidates. ProvenATS analyzes and understands them. We do not replace your ATS, we enhance it with psychographic intelligence that helps you make better hiring decisions.
Is candidate data secure?
Absolutely. We use AES-256-GCM encryption, never store raw PII, and hash candidate identifiers. Our architecture is SOC 2 ready, and we can sign custom BAA agreements for enterprise deployments.
Can I weight the dimensions differently for different roles?
Yes. Default tier weights ship out of the box and can be tuned per account in Settings. Each job description can also carry its own weighting override and per-dimension minimum targets (for example, "Conscientiousness at least 7"). The compare view marks every candidate cell green or red against those targets. Rankings recompute on read, so changing a weight updates results across the dashboard, library, and exports without re-running inference.
Can I supplement LinkedIn with a resume?
Yes. The analyze form accepts an optional pasted resume alongside the LinkedIn URL. It is stored AES-256-GCM encrypted and never returned to the browser; the candidate detail view simply marks "Resume captured." This is most useful for senior candidates whose LinkedIn footprints are sparse.
How accurate are the assessments?
Every dimension comes with a confidence score (high / medium / low) and 1-10 direct text citations from the candidate profile. Sparse profiles produce lower-confidence results, and Tier 1 detection of an active career transition automatically caps confidence across all downstream dimensions. Tier 1 structural analysis runs on Claude Opus and gates everything below it; Premium tier adds cross-framework convergence checks that flag contradictions for human review.
What do I actually receive when I buy?
You receive a complete white-label-ready Next.js application with full source code, deployed to your infrastructure or ours. This includes all features, unlimited users, managed hosting, security updates, and feature upgrades, forever. It is a one-time purchase, not a subscription.
What is the BYO Infrastructure option?
For Enterprise customers who want complete ownership, we offer a Bring Your Own Infrastructure add-on (+$10,000 one-time setup). We migrate the codebase to use your own accounts:
- - Your Postgres database and auth provider
- - Your AI inference API keys
- - Your LinkedIn data provider subscription
- - Your hosting account
You own everything. We provide the code, database migrations, and white-glove deployment. You pay the vendors directly (~$250-1,200/mo depending on volume). This is ideal for enterprises with strict data sovereignty or compliance requirements.
Ready to stop guessing?
Join recruitment agencies and enterprise teams who are replacing gut feelings with data-driven insights.
Or reach us at contact@provenlabs.ai
- 1. McFeely, S., & Wigert, B. (2019). This Fixable Problem Costs U.S. Businesses $1 Trillion. Gallup. Replacing an employee costs between one-half and two times their annual salary. gallup.com
- 2. Boushey, H., & Glynn, S. J. (2012). There Are Significant Business Costs to Replacing Employees. Center for American Progress. Meta-analysis of 30 case studies finding turnover costs reach up to 213% of salary for executive and specialized roles. americanprogress.org
- 3. Murphy, M. (2011). Hiring for Attitude. Leadership IQ. Three-year study of 5,247 hiring managers across 312 organizations finding that 46% of new hires fail within 18 months and 89% of those failures are attitude-driven, not skill-driven. leadershipiq.com
- 4. SHRM (2022). SHRM Benchmarking Report: $4,129 Average Cost-per-Hire. Subsequent SHRM 2024-2025 benchmarks place the figure near $4,700. shrm.org
- 5. Ashton, M. C., & Lee, K. (2007). Empirical, Theoretical, and Practical Advantages of the HEXACO Model of Personality Structure. Personality and Social Psychology Review, 11(2), 150-166. journals.sagepub.com
- 6. Schwartz, S. H. (1992). Universals in the Content and Structure of Values: Theoretical Advances and Empirical Tests in 20 Countries. Advances in Experimental Social Psychology, 25, 1-65. psycnet.apa.org
- 7. Holland, J. L. (1997). Making Vocational Choices: A Theory of Vocational Personalities and Work Environments (3rd ed.). Psychological Assessment Resources. psycnet.apa.org
- 8. Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2022). Revisiting Meta-Analytic Estimates of Validity in Personnel Selection: Addressing Systematic Overcorrection for Restriction of Range. Journal of Applied Psychology, 107(11), 2040-2068. doi.org/10.1037/apl0000994
- 9. Pletzer, J. L., Bentvelzen, M., Oostrom, J. K., & de Vries, R. E. (2019). A Meta-Analysis of the Relations Between Personality and Workplace Deviance: Big Five versus HEXACO. Journal of Vocational Behavior, 112, 369-383. doi.org/10.1016/j.jvb.2019.04.004