Job Summary

The Office of Responsible AI is seeking a professional with both AI engineering expertise and governance knowledge. The role involves reviewing agentic and generative AI solutions, assessing risks, validating controls, defining guardrails, and enabling teams to build responsible AI into production systems. The individual will translate technical findings into governance decisions that clients, delivery teams, and regulators can trust.

Key Responsibilities

Governance & Assurance

• Lead intake, risk classification, and governance reviews for AI and Agentic AI use cases.

• Own Responsible AI policies, controls, standards, and approval processes aligned to EU AI Act, NIST AI RMF, and ISO 42001.

• Support AI review boards, manage approvals, maintain AI inventories/model registries, and produce audit-ready documentation.

Technical Assurance

• Review agentic architectures, autonomy controls, human oversight mechanisms, and failure containment.

• Assess RAG pipelines for grounding quality, retrieval effectiveness, citation accuracy, and content leakage risks.

• Evaluate models for performance, drift, data provenance, licensing compliance, and business suitability.

• Define and validate evaluations covering accuracy, robustness, bias, toxicity, security, prompt injection, jailbreak resistance, and data protection.

• Establish guardrails, safety controls, policy enforcement, sandboxing, and adversarial testing practices.

Enablement & Advisory

• Advise architects and delivery teams on Responsible AI by design.

• Develop playbooks, training, reference patterns, and adoption guidance.

• Lead AI incident reviews and support client, audit, and regulatory engagements.

Skill Requirements

Agentic AI & Tooling

• Experience with LangGraph, LangChain, AutoGen, CrewAI, Semantic Kernel, OpenAI Agents SDK, MCP, function-calling, multi-agent patterns, and observability tools.

Models, Fine-Tuning & RAG

• Strong understanding of transformers, LLMs, tokenization, context management, and inference parameters.

• Experience with prompt engineering, instruction tuning, LoRA/QLoRA, fine-tuning, RLHF, and preference optimization.

• Hands-on experience with RAG architectures, embeddings, vector databases, retrieval optimization, and LLM evaluation frameworks.

• Understanding of deployment cost, latency, scalability, and sustainability trade-offs.

Engineering Foundations

• Proficiency in Python and AI application development.

• Experience with Azure AI Foundry, AWS Bedrock, or Google Vertex AI.

• Knowledge of MLOps/LLMOps, CI/CD, monitoring, versioning, security, privacy, and data governance.

AI Governance Expertise

• Knowledge of EU AI Act, NIST AI RMF, ISO 42001/23894, GDPR, DPDP Act, and AI risk management frameworks.

• Experience conducting AI impact assessments, vendor reviews, assurance activities, red teaming, explainability, fairness, transparency, and accountability assessments.

• Familiarity with regulated industry requirements and model risk management practices.

Behavioral Competencies

• Strong stakeholder management, influence without authority, and executive communication skills.

• Sound judgment, constructive challenge, ethical decision-making, and risk-based thinking.

• Effective collaboration across engineering, security, legal, privacy, procurement, HR, and business teams.

• Strong facilitation, coaching, problem-solving, and continuous learning mindset.

Qualifications & Experience

• Bachelor’s/Master’s degree in Computer Science, Engineering, Data Science, or related field.

• 8-12 years of technology experience, including 3+ years in ML/Generative AI.

• 2+ years in AI Governance, Model Risk, AI Assurance, or Responsible AI roles.

• Experience delivering governed Generative AI or Agentic AI solutions into production.

• Client-facing, consulting, audit, or regulatory engagement experience preferred.

Preferred Certifications

• IAPP AIGP.

• ISO 42001 Lead Implementer/Auditor.

• Azure AI Engineer, AWS ML Specialty, or Google Professional ML Engineer.

• CISSP, CIPP/E, CIPM, or equivalent.

• Publications, open-source contributions, or conference presentations on Responsible AI.

Other Requirements

First-Year Success Measures

• All AI use cases classified, inventoried, and mapped to controls with assigned ownership.

• Consistent AI governance review process adopted across delivery teams.

• Automated evaluation and red-team testing integrated into production AI workloads.

• Delivery teams self-serve governance requirements with minimal escalation.