
EY - GDS Consulting - AI and DATA - Responsible AI - Senior
- Bangalore, Karnataka
- Permanent
- Full-time
- Establish and run LLMOps/Agentic Ops practices (lifecycle, approvals, versioning, observability, incident response/playbooks) integrated with platform tooling.
- Define and enforce governance & security controls (PII protection, data residency, model access, content safety, jailbreak/prompt injection defenses), integrating with enterprise security.
- Build LLM evaluation pipelines (groundedness/faithfulness, toxicity/PII, bias/fairness, robustness) and quality gates for pre prod and ongoing post deployment monitoring.
- Implement guardrails/ethics (policy as code, allow/deny lists, safety filters, red teaming harnesses) and ensure compliant documentation (model cards, data sheets, DPIAs).
- Contribute production grade Python libraries, policies, and APIs that product teams can adopt “as a service”; partner with platform teams on AWS/Azure controls.
- LLMOps/Agentic Ops patterns and tooling
- Governance/Security (threat modeling for LLMs, privacy, access controls)
- LLM Evaluation frameworks (human+automatic metrics; eval harnesses)
- Guardrails/Ethics techniques and incident playbooks
- Python
- AWS/Azure
- API development for governance services
- Familiarity with RAG/advanced agentic AI to guide safe design choices
- B.Tech/M.Tech/MS in CS/EE or equivalent.
- 4+ years in Responsible AI Engineering