AI ethics at Tarkflo
A real conversation.
A human decision.
Role-relevant conversations. Evidence your team can review.
Tarkflo connects interview evidence, clear evaluation criteria and human judgement. Your team shapes the process, chooses review checkpoints and makes informed hiring decisions.
Understand the approachIndia-first overview · Reviewed 5 November 2026
01 / The interview logic
Ask about the work.
Listen for the experience.
Build the conversation around relevant experience: what a person did, how they approached decisions and what they learned. Follow-ups help your team understand the reasoning behind the response.
Start with the role.
The hiring team sets the role, level, focus areas and evaluation criteria. Job-description and CV context help shape the conversation when enabled and available.
Explore relevant experience.
Questions explore concrete work, personal actions, decisions and outcomes. Role-relevant scenarios and knowledge checks give candidates further ways to demonstrate their understanding.
Build on each answer.
Follow-ups explore examples, reasoning, trade-offs and validation from the conversation. The interviewer is guided to distinguish demonstrated experience from hypothetical reasoning.
Connect evidence with judgement.
AI-assisted evaluations and supporting responses give the hiring team a shared basis for review. Your team sets the checkpoints, considers the candidate’s context and owns the hiring decision.
An illustrative question pattern
Experience.
Reasoning. Context.
These are explanatory examples, not a real candidate transcript or a fixed script.
“What is stakeholder management?”
“Tell me about a project where two stakeholders wanted different outcomes. What did you personally do?”
“What trade-off did you make, and how did you know it worked?”
02 / Evidence and evaluation
Evidence to explore.
Judgement for your team.
AI organises response evidence and generates evaluations against configured criteria. Your team defines the progression rules, reviews the supporting material and remains responsible for the hiring decision.
How the responsibilities differ
Source responses
What the candidate said, with available supporting material.
AI-assisted evaluation
Scores and summaries connected to configured criteria.
Human review
Consider the evidence, the role and the candidate’s context.
Human review and automation follow the checkpoints your team configures.
AI-assisted work
Make evidence reviewable.
Bring role-based evaluations, supporting answers and available transcripts or recordings into the candidate review. Use that shared evidence to prepare a focused follow-up conversation.
Human responsibility
Lead the review and decision.
Define the criteria, review checkpoints and progression rules, including any automated stage transitions. Consider candidate context and accessibility needs, document your review and decide what follows.
03 / Controls for your process
Clear controls.
Considered decisions.
Configure review checkpoints, integrity policies and evidence sharing around your hiring process. Pair those controls with clear candidate information and accountable team review.
Give candidates clear information.
Candidate notices explain AI interviewing and the recording, scoring and integrity processing configured for the round. The hiring organisation defines the purpose, supplies appropriate notices and obtains any required consent.
Keep progression in your team’s control.
Configure invitation rules and review checkpoints for your process. Applications below a CV FitScore gate are held with a pending rejection proposal that an authorised reviewer can confirm or dismiss. Qualifying applications and other stages can advance under configured rules.
Choose a clear integrity policy.
Configure camera, face-comparison, device or voice checks where enabled, separately from answer evaluation. Review integrity signals in context: a flag is a review aid, not proof of misconduct. Configured enforcement can pause or end a session under the organisation’s chosen policy.
Share the right evidence.
Select report fields, set expiry and revoke shared access. Share public report links only with intended reviewers: anyone who holds a valid link can access the selected content until it expires or is revoked.
Manage candidate data purposefully.
Owner-controlled candidate erasure and retention settings for raw integrity event logs support your data-management process. Set appropriate policies for candidate records, recordings and reports, and provide a clear route for data requests.
Assess job-relevant content.
Answer-grading instructions focus on response content and exclude protected-attribute inference, accent and appearance penalties. Evaluation distinguishes assessed evidence from skills still to explore. Separate communication insights support review of role-relevant communication.
04 / India-first regulatory context
Responsible hiring.
A practical approach.
Build responsible hiring around appropriate job criteria, clear candidate communication and purposeful data management. The hiring organisation remains responsible for its deployment and its use of AI-assisted results.
This product overview supports deployment planning. Consult a qualified adviser for legal guidance on the requirements that apply to your organisation.
Clear purpose and candidate communication.
India’s DPDP framework provides for consent and specified legitimate uses. Hiring organisations should identify the appropriate basis for their circumstances, explain candidate-data use and establish access, retention and candidate-request processes.
A practical DPDP implementation plan.
The DPDP Act and final Rules have phased commencement. Use the official provisions and applicable dates to plan your notices, data-management processes and candidate-request handling, with deployment-specific legal guidance.
Human-centric AI in everyday practice.
India’s AI Governance Guidelines promote responsible, risk-based and evidence-led deployment. Put those principles into practice through job-relevant criteria, clear candidate information, meaningful review and accountable use of AI-assisted results.
For candidates
Your context deserves
to be heard.
For accessibility or language support, questions about your responses, or additional context you would like considered, contact the hiring organisation that invited you. Share any technical issue or transcript correction, and ask for human review or an appropriate alternative assessment.
For recruiting-data requests, start with that organisation. Our privacy policy explains Tarkflo’s role and request channels.
Clear answers about AI in hiring.
Role-relevant conversations. Considered decisions.
See responsible hiring in action.
Explore how role-specific interviews, reviewable evidence and configurable controls work together in your hiring process.
A product walkthrough built around your team’s process.