Job Description
The Responsible AI Analyst supports EXL’s Responsible AI program by evaluating AI/GenAI solutions for governance, fairness, transparency, security/privacy, hallucination risk, and regulatory readiness. The analyst will work closely with delivery teams, AI engineers, InfoSec, Data Privacy, and business stakeholders to identify Responsible AI gaps, recommend mitigations, and produce audit-ready documentation.
Key Responsibilities
1) Responsible AI Risk Assessments
- Conduct Responsible AI evaluations for AI/ML/GenAI solutions across EXL projects.
- Perform gap assessments across:
- Model documentation & transparency
- Human-in-the-loop (HITL) oversight
- Bias & fairness
- Hallucination & reliability
- Monitoring, drift & retraining readiness
- Explainability & traceability
- Content safety & toxicity controls
- Data quality/integrity validation
- Security & privacy controls
- Environmental sustainability (where applicable)
2) Bias & Fairness Evaluations
- Execute bias assessments using fairness frameworks and metrics such as:
- Disparate Impact (DI)
- Demographic Parity
- Equal Opportunity
- FPR/FNR parity
- Review protected category risks and proxy bias risks (region, income proxy, etc.).
- Document fairness findings and mitigation proposals.
3) GenAI Specific Controls
- Evaluate hallucination controls and reliability measures:
- grounding (RAG / citations)
- golden datasets
- confidence scoring
- human review triggers
- Review content safety controls:
- toxic content filters
- prompt injection/jailbreak prevention
- guardrails policies
4) Governance & Documentation
- Prepare regulator-ready and audit-ready documentation:
- Responsible AI Evaluation Report
- AI Bias Evaluation Report
- Executive risk summary dashboards
- Work notes / approval summaries
- Maintain standard templates and control checklists.
- Support model registration processes on ServiceNow.
5) Stakeholder Collaboration
- Engage with BU teams to define mitigation plans and approval conditions.
- Coordinate with InfoSec and Data Privacy teams for security/privacy sign-offs.
- Support AI Ops Committee / governance forums with evidence and analysis.
Key Responsibilities
1) Responsible AI Risk Assessments
- Conduct Responsible AI evaluations for AI/ML/GenAI solutions across EXL projects.
- Perform gap assessments across:
- Model documentation & transparency
- Human-in-the-loop (HITL) oversight
- Bias & fairness
- Hallucination & reliability
- Monitoring, drift & retraining readiness
- Explainability & traceability
- Content safety & toxicity controls
- Data quality/integrity validation
- Security & privacy controls
- Environmental sustainability (where applicable)
2) Bias & Fairness Evaluations
- Execute bias assessments using fairness frameworks and metrics such as:
- Disparate Impact (DI)
- Demographic Parity
- Equal Opportunity
- FPR/FNR parity
- Review protected category risks and proxy bias risks (region, income proxy, etc.).
- Document fairness findings and mitigation proposals.
3) GenAI Specific Controls
- Evaluate hallucination controls and reliability measures:
- grounding (RAG / citations)
- golden datasets
- confidence scoring
- human review triggers
- Review content safety controls:
- toxic content filters
- prompt injection/jailbreak prevention
- guardrails policies
4) Governance & Documentation
- Prepare regulator-ready and audit-ready documentation:
- Responsible AI Evaluation Report
- AI Bias Evaluation Report
- Executive risk summary dashboards
- Work notes / approval summaries
- Maintain standard templates and control checklists.
- Support model registration processes on ServiceNow.
5) Stakeholder Collaboration
- Engage with BU teams to define mitigation plans and approval conditions.
- Coordinate with InfoSec and Data Privacy teams for security/privacy sign-offs.
- Support AI Ops Committee / governance forums with evidence and analysis.
The Responsible AI Analyst supports EXL’s Responsible AI program by evaluating AI/GenAI solutions for governance, fairness, transparency, security/privacy, hallucination risk, and regulatory readiness. The analyst will work closely with delivery teams, AI engineers, InfoSec, Data Privacy, and business stakeholders to identify Responsible AI gaps, recommend mitigations, and produce audit-ready documentation.
Key Responsibilities
1) Responsible AI Risk Assessments
- Conduct Responsible AI evaluations for AI/ML/GenAI solutions across EXL projects.
- Perform gap assessments across:
- Model documentation & transparency
- Human-in-the-loop (HITL) oversight
- Bias & fairness
- Hallucination & reliability
- Monitoring, drift & retraining readiness
- Explainability & traceability
- Content safety & toxicity controls
- Data quality/integrity validation
- Security & privacy controls
- Environmental sustainability (where applicable)
2) Bias & Fairness Evaluations
- Execute bias assessments using fairness frameworks and metrics such as:
- Disparate Impact (DI)
- Demographic Parity
- Equal Opportunity
- FPR/FNR parity
- Review protected category risks and proxy bias risks (region, income proxy, etc.).
- Document fairness findings and mitigation proposals.
3) GenAI Specific Controls
- Evaluate hallucination controls and reliability measures:
- grounding (RAG / citations)
- golden datasets
- confidence scoring
- human review triggers
- Review content safety controls:
- toxic content filters
- prompt injection/jailbreak prevention
- guardrails policies
4) Governance & Documentation
- Prepare regulator-ready and audit-ready documentation:
- Responsible AI Evaluation Report
- AI Bias Evaluation Report
- Executive risk summary dashboards
- Work notes / approval summaries
- Maintain standard templates and control checklists.
- Support model registration processes on ServiceNow.
5) Stakeholder Collaboration
- Engage with BU teams to define mitigation plans and approval conditions.
- Coordinate with InfoSec and Data Privacy teams for security/privacy sign-offs.
- Support AI Ops Committee / governance forums with evidence and analysis.
Key Responsibilities
1) Responsible AI Risk Assessments
- Conduct Responsible AI evaluations for AI/ML/GenAI solutions across EXL projects.
- Perform gap assessments across:
- Model documentation & transparency
- Human-in-the-loop (HITL) oversight
- Bias & fairness
- Hallucination & reliability
- Monitoring, drift & retraining readiness
- Explainability & traceability
- Content safety & toxicity controls
- Data quality/integrity validation
- Security & privacy controls
- Environmental sustainability (where applicable)
2) Bias & Fairness Evaluations
- Execute bias assessments using fairness frameworks and metrics such as:
- Disparate Impact (DI)
- Demographic Parity
- Equal Opportunity
- FPR/FNR parity
- Review protected category risks and proxy bias risks (region, income proxy, etc.).
- Document fairness findings and mitigation proposals.
3) GenAI Specific Controls
- Evaluate hallucination controls and reliability measures:
- grounding (RAG / citations)
- golden datasets
- confidence scoring
- human review triggers
- Review content safety controls:
- toxic content filters
- prompt injection/jailbreak prevention
- guardrails policies
4) Governance & Documentation
- Prepare regulator-ready and audit-ready documentation:
- Responsible AI Evaluation Report
- AI Bias Evaluation Report
- Executive risk summary dashboards
- Work notes / approval summaries
- Maintain standard templates and control checklists.
- Support model registration processes on ServiceNow.
5) Stakeholder Collaboration
- Engage with BU teams to define mitigation plans and approval conditions.
- Coordinate with InfoSec and Data Privacy teams for security/privacy sign-offs.
- Support AI Ops Committee / governance forums with evidence and analysis.
Graduate in Science, Professional graduate in AI and ML, computer science, information science.
