Tenzo AI
Tenzo AI's AI recruiter, Morgan, screens and interviews applicants by phone, SMS, email, WhatsApp or video in 40+ languages, scores them against employer-defined rubrics with fraud and identity checks, schedules next rounds, and writes scores, transcripts and recordings back to the customer's ATS.
HireAIScore rates Tenzo AI 56 out of 100 (grade F, Substantial gaps) under rubric v1.0, last reviewed September 25, 2026. Tenzo AI ranks 23 of 160 vendors rated in Screening AI, against a category average of 46. That score is drawn from 34 evidence items across 7 rubric criteria for Tenzo AI — 30 with a cited source, 4 recording that nothing was located.
§ 01 - Company facts
- Legal name
- Salv AI, Inc.
- Founded
- 2024
- Headquarters
- United States · US
- Pricing tier
- Mid
- Market side
- Employer-side (AEDT)
- Categories
- ScreeningVideo interviewScheduling
- Website
- tenzo.ai
These are company details, not findings. They are editable by a verified representative and carry no weight in the score, which is built only from the cited evidence below.
§ 02 - Score breakdown
§ Score breakdown
Category scoring
Weighted contribution shown to the right of each bar.
- 01
Article 11 Technical Documentation
Weight 20%45
+9.0 · category avg 49
- 02
Bias Audit Transparency
Weight 18%83
+14.9 · category avg 41
- 03
FRIA Support
Weight 15%40
+6.0 · category avg 31
- 04
Data Governance Disclosure
Weight 15%55
+8.3 · category avg 53
- 05
Human Oversight Design
Weight 12%60
+7.2 · category avg 56
- 06
Post-Market Monitoring
Weight 12%45
+5.4 · category avg 40
- 07
Customer Documentation
Weight 8%62
+5.0 · category avg 58
Category avg is the mean raw score on that criterion across the 160 Screening AI vendors in scope of this rubric, this one included.
§ 03 - Strongest · weakest
Strongest category
Bias Audit Transparency
Raw score 83 · contributes 14.9 to total.
83 against a 41 category average
Weakest category
FRIA Support
Raw score 40 · contributes 6.0 to total.
40 against a 31 category average
§ 04 - Cited evidence
Download diligence record→§ Evidence
Cited per category
Every score is backed by at least one cited piece of evidence.
Evidence ledger
- Items
- 34
- Documentation
- 19
- Audit report
- 3
- Public statement
- 8
- Absence
- 4
- With a source URL
- 30 of 34
- Source hostnames
- 4
- Fewest items
- 4
- FRIA Support
These figures measure how thoroughly Tenzo AI was reviewed, not how Tenzo AI performed — an absence row, recording that nothing was located, is counted like any other item.
Article 11 Technical Documentation
5 items45
- DocumentationCaptured Sep 25, 2026
Warden AI's published report describes the AI Screening Agent's inputs (interview questions, grading criteria, candidate responses) and its single output, a match score, which is the closest thing to a public system description.
- Public statementCaptured Sep 25, 2026
Tenzo says it does not claim conformity with the AI Act's high-risk regime, that documentation and logging work toward the December 2027 date is in progress, and that multiple models run in parallel for redundancy.
https://www.tenzo.ai/blog/ai-interviewing-compliance-eeoc-eu-ai-act-buyers-guide
- Public statementCaptured Sep 25, 2026
An October 2025 post explains Tenzo's design principle: LLMs score each answer on its own against job-relevant criteria instead of ranking résumés, and humans evaluate the evidence.
https://www.tenzo.ai/blog/moving-beyond-black-box-fears-new-questions-for-ai-in-hiring
- DocumentationCaptured Sep 25, 2026
Platform Agreement §4(b) commits Tenzo to give customers model documentation and bias audit results on written request at no extra cost, so the documentation exists but is not public.
- AbsenceCaptured Sep 25, 2026
No model or system card, explainability statement, instructions for use or ISO 42001 certificate was found on tenzo.ai, its blog, its legal pages or trust.tenzo.ai, and docs.tenzo.ai requires a login.
Bias Audit Transparency
5 items83
- Audit reportCaptured Sep 25, 2026
Warden AI's public dashboard shows monthly independent audits of Tenzo's AI Screening Agent. The latest, on 24 September 2026, used a 44,992 sample and returned 110 Clear, 0 Consider and 0 Concern results across 17 bias categories including sex × race, plus a zip-code sensitivity test.
- Audit reportCaptured Sep 25, 2026
The NYC LL 144 report dated 24 September 2026 publishes impact ratios (sex 0.99, race/ethnicity 0.96–1.00, intersectional 0.93–1.00), but it runs on Warden's test dataset 'due to a lack of access to historical data'.
- DocumentationCaptured Sep 25, 2026
Warden's case study says Tenzo chose monthly bias audits with disparate-impact and counterfactual testing so it can catch bias before model updates and feature launches.
https://www.warden-ai.com/case-studies/how-tenzo-maps-ai-interviewing-to-californias-feha-rules
- Public statementCaptured Sep 25, 2026
Tenzo's July 2026 post says the platform undergoes monthly third-party bias audits but neither names the auditor nor links the public dashboard.
https://www.tenzo.ai/blog/building-an-ai-recruiter-in-house-vs-tenzo
- AbsenceCaptured Sep 25, 2026
None of the tenzo.ai pages, blog posts or trust.tenzo.ai link to the Warden dashboard at trust.warden-ai.com; only one blog post links Warden's case study, and the Platform Agreement offers audit results on request.
FRIA Support
4 items40
- DocumentationCaptured Sep 25, 2026
Tenzo's February 2026 compliance guide explains EU AI Act deployer duties, such as following the instructions for use and keeping logs for at least six months, and notes that Article 27 FRIAs apply only to certain deployers.
- DocumentationCaptured Sep 25, 2026
Platform Agreement §4(b) makes the customer, as deployer, responsible for impact assessments, notices, consents and human review, and has Tenzo supply model documentation and bias audit results on request.
- DocumentationCaptured Sep 25, 2026
The public DPA commits Tenzo to reasonable help with data protection impact assessments and prior consultation, with the customer paying for any non-negligible effort.
- AbsenceCaptured Sep 25, 2026
No Fundamental Rights Impact Assessment template, worksheet or Article 27 pack for deployers was found on tenzo.ai, its blog, its legal pages or trust.tenzo.ai.
Data Governance Disclosure
5 items55
- DocumentationCaptured Sep 25, 2026
The privacy policy, updated 27 August 2026, lists the candidate data processed: recordings, transcripts, scoring outputs, ID data and, where enabled, biometric data. It leaves retention to the employer and requires biometric data to be destroyed within three years of the last interaction.
- DocumentationCaptured Sep 25, 2026
The public DPA gives customers 10 days to object to new subprocessors, deletes data at termination and uses SCCs, but supplies the subprocessor list only on request and lets Tenzo use customer personal data to build or improve its products.
- Public statementCaptured Sep 25, 2026
The homepage says a de-biasing layer removes protected-class information from transcripts before evaluation and that the platform is SOC 2 Type 2 and GDPR compliant.
- Public statementCaptured Sep 25, 2026
Tenzo says it evaluates candidates only on information they submit and data generated in the hiring workflow, and does not scrape the public web to build hidden candidate profiles.
- DocumentationCaptured Sep 25, 2026
Tenzo's trust center lists SOC 2 Type II and Type I compliance with request-gated reports, plus controls for data retention and deletion, customer data deletion, and encryption at rest and in transit.
Human Oversight Design
6 items60
- DocumentationCaptured Sep 25, 2026
Scorecards can be customized per requisition (criteria, weighting, knockouts) and show per-criterion and overall scores with transcript, recording and ID-verified flags; Tenzo says Morgan does not make the hiring decision and a human decides.
- DocumentationCaptured Sep 25, 2026
The Solutions FAQ says a person reads Morgan's rubric-scored, recorded and transcribed record and makes the decision, and that everyone on a requisition is scored against the same employer-defined rubric.
- Public statementCaptured Sep 25, 2026
Tenzo describes recruiter review of full transcripts, version history for audit and documented accommodation paths, and says humans review throughout and make all final decisions.
https://www.tenzo.ai/blog/ai-interviewing-compliance-eeoc-eu-ai-act-buyers-guide
- Public statementCaptured Sep 25, 2026
The homepage shows a compliance rule, marked as always applied, that flags asking candidates their current salary as a violation, and says candidates can opt out of AI interviews and request accommodations.
- Public statementCaptured Sep 25, 2026
Fraud and identity signals, such as face-match results, tab switching and more than one person on camera, go to recruiters for review, and Tenzo says they should inform human decisions, not replace them.
https://www.tenzo.ai/blog/candidate-fraud-deepfake-prevention
- DocumentationCaptured Sep 25, 2026
Platform Agreement §8(b) requires the customer to review all AI-generated materials, including candidate evaluations and scores.
Post-Market Monitoring
4 items45
- Audit reportCaptured Sep 25, 2026
The Warden AI dashboard provides ongoing post-market bias monitoring: it re-audits Tenzo's screening agent every month and publishes the latest results, most recently for 24 September 2026.
- DocumentationCaptured Sep 25, 2026
The trust center lists incident response, security incident logging and customer notification of major changes among its controls, but its Updates feed shows 'No updates yet' and the Incident Response Policy is request-gated.
- DocumentationCaptured Sep 25, 2026
The DPA commits Tenzo to give written notice of security incidents without undue delay and within legally required time frames, with no fixed number of hours.
- AbsenceCaptured Sep 25, 2026
Tenzo has no public status page, security.txt or changelog of model and rubric changes. status.tenzo.ai does not resolve; tenzo.ai/.well-known/security.txt, /status and /changelog return 404; and the pages at tenzo.statuspage.io and tenzo.betteruptime.com cannot be tied to Tenzo AI.
Customer Documentation
5 items62
- DocumentationCaptured Sep 25, 2026
The solutions hub has 16 pages by industry and hiring motion, with FAQs on how Morgan scores, how humans decide, what is written back to the ATS and which compliance certifications Tenzo holds.
- DocumentationCaptured Sep 25, 2026
Candidate terms, updated 21 February 2026, state that the employer makes all hiring decisions. They include an EU AI Act transparency section (notice of AI interaction, right to explanation), a BIPA notice and consent, and an EU GDPR representative.
- DocumentationCaptured Sep 25, 2026
The full customer Platform Agreement is public, including its AI disclaimers, the customer's duties under laws on automated decision tools, and availability and SLA terms.
- DocumentationCaptured Sep 25, 2026
The candidate FAQ explains how to request accommodations (on the call, by SMS, by email or in the video portal) and how to reschedule, rejoin or fix technical problems.
- DocumentationCaptured Sep 25, 2026
Tenzo's current compliance guide for buyers covers EU AI Act Article 5(1)(f), Article 50 and the Annex III dates, EEOC adverse-impact analysis, NYC LL 144, Illinois HB 3773, Texas, California FEHA and Colorado SB 26-189.
https://www.tenzo.ai/blog/ai-interviewing-compliance-eeoc-eu-ai-act-buyers-guide
§ 05 - Editorial notes
Company overview
Tenzo AI is the trading name of Salv AI, Inc., a Delaware corporation headquartered at 425 California St in San Francisco. Its LinkedIn page gives a 2024 founding date and 11–50 staff. Its AI recruiter, Morgan, contacts applicants within minutes by phone, text, email, WhatsApp or video and interviews them in more than 40 languages. It then writes a rubric-based score, a full transcript and the recording back to more than 57 ATS and CRM platforms. Other agents handle résumé review, fraud and identity checks, scheduling and rediscovery of past applicants. Customers named on its site include staffing firms such as Agilus Work Solutions, Synergie and Game 7 Staffing, and employers such as Sequoia Living and Cascade Environmental. Pricing is by quote only.
Regulatory exposure
Morgan's rubric-based match scores and ranked shortlists shape who advances. NYC employers that use them to substantially assist hiring decisions therefore need Local Law 144's annual independent bias audit, a published summary and candidate notice. Tenzo's monthly Warden AI audit covers the audit itself, but notice and publication stay with the employer. In the EU, Tenzo is a high-risk system under Annex III point 4. Those obligations were deferred to 2 December 2027, but the Article 5(1)(f) ban on emotion recognition in the workplace already applies, and Tenzo's candidate terms still provide for employer-deployed "emotion recognition or biometric categorization features". Recorded video interviews and optional face-match ID checks bring in Illinois's AI Video Interview Act and BIPA; for BIPA, Tenzo publishes a consent form and a three-year destruction schedule. Illinois HB 3773's notice duty has applied since 1 January 2026. California's FEHA rules on automated decisions require records to be kept for four years, and Colorado's SB 26-189 disclosure regime starts on 1 January 2027. Tenzo's Platform Agreement makes the customer responsible for notices, consents, impact assessments and human review.
Path to a higher score
Tenzo's bias testing is strong. What holds its score down is thin documentation that is hard to find. A public system card or Responsible AI page would help most: which models it uses, what the de-biasing layer removes, how match scores are built, known limits and instructions for use. That page should carry the model documentation the contract already offers on request, and a list of the fields Morgan infers that confirms it draws no emotional inferences. Tenzo should link the Warden dashboard from tenzo.ai and the trust center, and audit production data as well as test data. It should publish its subprocessor and model-provider list, say whether customer data trains its models, and give customers NYC, Illinois and Colorado notice templates plus an Article 27 FRIA template. For monitoring, it needs a status page, a changelog of model and rubric changes, and a security.txt. Its February 2026 compliance guide still gives dates for Colorado's repealed SB 24-205 and should be corrected.
§ Regulatory frame
What applies to screening ai.
Treated as high-risk under Annex III §4 when the screening output is decisive. EU AI Act Article 13 transparency obligations apply; deployers must give candidates a way to contest.
§ Compare
Build any comparison→Tenzo AI against its nearest-scoring peers.
In Screening AI.
§ Others rated in Screening AI
All screening vendors→Ranked 23 of 160 by weighted total under rubric v1.0. The ordering is arithmetic on the rubric and carries no view on which tool suits a given hiring process.
- 20Ashby56F
- 21Tribepad56F
- 22iMocha56F
- 23Tenzo AIThis profile56F
- 24HackerRank55F
- 25Gem55F
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.