Zeenna AI system instructions for use

1. Intended purpose

Zeenna is a decision-support tool for the initial pre-screening stage of recruitment. It conducts structured text interviews, collects candidate responses and prepares clear summaries for recruiters. It supports human decision-making and does not replace it.

2. Permitted and prohibited use

• Permitted: candidate pre-screening within an agreed workflow.
• Permitted: response summaries, approved follow-up questions and supportive evaluation against deployer-approved rubrics.
• Not permitted: making hiring or rejection decisions without human assessment.
• Not permitted: using Zeenna as the sole source for a negative decision.
• Not permitted: evaluating candidates using protected or discriminatory characteristics.
• Not permitted: transferring responsibility for the hiring decision from the deployer to the AI system.

3. System limitations

The quality of generated questions and scoring depends on the quality and detail of the job description, knowledge base and evaluation rubrics configured by the deployer.

Zeenna does not extract biometric data for identification, infer emotions from voice, or profile personality traits. Voice input is used only to transcribe or record a candidate response.

Large language models can produce inaccurate outputs. Zeenna reduces this risk through approved knowledge sources, a human fallback and mandatory human oversight.

4. Robustness, cybersecurity and change management

Zeenna operates in an environment designed for security, availability, data protection and controlled change. Measures include access management, MFA, restricted production-database access, security monitoring, backups, vulnerability management and permission reviews.

Changes that may materially affect integrations, security, availability or data processing are communicated through the appropriate contractual, operational or technical documentation.

5. Deployer responsibilities

• Ensure trained and competent people operate the system.
• Review the supporting report before making a decision with legal or similarly significant effect.
• Allow authorised reviewers to disregard the system score and decide differently.
• Inform candidates before screening that they will interact with an AI system.
• Do not place discriminatory requirements or special-category personal data into rubrics, knowledge sources or prompts unless expressly required and permitted by law.

6. Human oversight controls

• Recruiters can review and manually edit generated questions before activation.
• Recruiters can pause or stop the pre-screening process for a position.
• AI-generated follow-up questions, reports and summaries are visibly labelled.
• Underlying candidate responses remain available separately from AI summaries.
• Unknown candidate questions are escalated to a recruiter rather than answered without a reliable source.

7. Logging and traceability

The administrative interface provides access to relevant process records. Depending on the workflow, these can include configuration, approved questions, rubric changes, candidate responses, generated follow-up questions, summaries, scores, candidate questions and human actions.

8. Summary

Zeenna remains a supporting tool. The authorised deployer is responsible for the final recruitment decision. Human oversight, candidate transparency and traceability must remain active throughout use.