| Company | Deloitte Touche Tohmatsu India LLP |
| Job Title | T&T | Cyber CST | Deputy Manager | AI Risk Management |
| Job Requisition ID | 111662 |
| Location | Bengaluru, Karnataka, India |
| Job Level | Deputy Manager |
| Job Function | Cyber / AI Risk Management |
| Role Focus | Enterprise-grade AI Engineering, Generative AI, Agentic AI, RAG/GraphRAG and Machine Learning |
| Core Requirement | Hands-on experience in designing, building and productionizing enterprise AI solutions |
| Programming | Strong fundamentals in Python |
| AI/ML Skills | ML/LLM concepts, retrieval architectures, agent orchestration, backend engineering, cloud deployment and AI evaluation |
| LLM Development | Build applications using commercial and open-source LLMs |
| LLM Workflows | Prompt engineering, structured outputs, tool/function calling and model routing |
| LLM Fundamentals | Tokenization, embeddings, attention, context windows, temperature, top-p/top-k sampling and token generation |
| Model Evaluation | Evaluate models based on quality, latency, throughput, concurrency, context limits and cost |
| Model Strategy | Design model abstraction and fallback strategies to reduce dependency on a single LLM provider |
| Optimization | Optimize AI applications for latency, token utilization and inference cost |
| Agentic AI | Design single-agent and multi-agent architectures for complex enterprise workflows |
| Architecture Decision | Determine when to use Agentic AI versus deterministic orchestration or traditional software workflows |
| Frameworks | LangGraph or equivalent agent orchestration frameworks |
| Engineering Requirements | Design systems for reliability, latency, concurrency, security, scalability, cost efficiency and failure recovery |
| Educational Qualification | Any Graduate / BE / B.Tech from Tier 1 or Tier 2 institute in CS / IT / EC / E, OR MCA |
| University Requirement | Recognized university |
