AI skills, demonstrated
See the work behind the credential.
Certavant credentials belong to the people who earn them. Individuals demonstrate their skills; companies support their own teams' development.
Performance and evidence
Assessments use realistic work products. Rubric criteria and cited evidence connect the result to what you demonstrated.
Human credential decisions
AI supports scoring. A named Certification Decision Maker makes the credential decision and can accept, modify, or reject the recommendation.
How grading works
- Your work is prepared for assessment.
- AI-assisted grading uses a fixed, versioned configuration and independent scoring passes, with evidence tied to the submission.
- Uncertainty, disagreements, borderline results, failures, perfect scores, and integrity flags receive the required human review.
- A Certification Decision Maker records the final decision. An uninvolved human reviewer handles appeals.
Deterministic scenario choices can receive rule-based feedback. AI-generated feedback is labelled where it appears. Read the AI transparency explanation.
Assessment blueprints describe the task families and scoring approach. Inspect the blueprints.
Your evidence, your visibility settings
Credential verification pages show the current status, issue and expiry dates, and the evidence available to that viewer. You choose public, employer-link-only, or hidden visibility for each artifact in your wallet.
Employer links are scoped and expire. Hidden artifacts are withheld from verification views. The credential's status remains independently verifiable.
Data handling
Learning, assessment, and credential records are stored in Certavant's platform. Assessment audit trails record grading configurations, evidence, and human decisions. Access to company and learner records is restricted to authorized users.
AI features send the relevant prompt, lesson context, conversation, or submitted work through Certavant's backend to the selected AI provider. Use the material provided in tasks and avoid entering confidential or personal data into AI interactions.
Payment processing is handled by Stripe. Published evidence and scoped employer links are intended to be shared with the viewers you choose.
Service providers and subprocessors
The configured provider depends on the feature and its approved routing. These services support the platform:
| Provider | Purpose | Relevant information |
|---|---|---|
| Anthropic | AI processing and assessment support | Relevant prompts, context, and submitted work for the selected AI task |
| OpenRouter | AI request routing and processing | Relevant request content sent to the selected model route |
| NUX | Configured AI processing endpoint | Relevant request content for approved NUX routes |
| Stripe | Payments and billing | Checkout and billing information |
| Cloudflare | Delivery and perimeter services | Web requests and connection information |