Join Deloitte's Technology & Transformation practice as a Manager | Backend Developer in Mumbai. This role is central to our mission of unlocking data's potential through innovative technology solutions. You'll be instrumental in leveraging AI, GenAI, and autonomous agents to build cutting-edge enterprise applications.

Our team specializes in strategic guidance and implementation services, helping companies manage data effectively and transform it into actionable insights. We work across business intelligence, data management, performance management, and next-generation analytics, including big data, cloud, and machine learning.

This is an exciting opportunity to lead the development of advanced agentic applications and contribute to the evolution of data-driven decision-making.

Lead the architecture, design, and development of sophisticated agentic applications, harnessing AI, GenAI, and autonomous agent paradigms for diverse enterprise use cases.

Ensure end-to-end technical delivery within Agentic Pods, adhering to rigorous standards for scalability, resilience, security, and production readiness.

Architect intelligent systems that fluidly integrate LLMs, tools, APIs, data pipelines, and enterprise workflows.

Drive the seamless integration of agentic platforms with existing enterprise applications, APIs, data platforms, and cloud services.

Define and implement robust design patterns, reusable components, accelerators, and reference architectures for agent-based systems.

Champion Responsible AI principles, embedding governance, explainability, guardrails, risk controls, and compliance into all solution designs.

Collaborate closely with business, product, and architecture stakeholders to translate complex requirements into scalable, agent-driven solutions.

Continuously optimize agent performance, accuracy, latency, cost efficiency, and system throughput.

Lead production troubleshooting, root cause analysis, and system stabilization to guarantee SLA adherence and operational excellence.

Provide technical mentorship and conduct design/code reviews for junior consultants, fostering overall engineering maturity.

Actively contribute to Agentic Pod standards, innovation initiatives, accelerators, and thought leadership within the team.

A Bachelor of Engineering (BE) or Bachelor of Technology (B.Tech) in Computer Science, Engineering, or a related field is required.

You should possess 7–9 years of substantial experience in designing and delivering agentic applications, utilizing AI, GenAI, and autonomous agent architectures.

Demonstrated expertise in enterprise-scale agentic system design, encompassing multi-agent coordination, autonomous workflows, and decision-making systems, is essential.

Strong hands-on proficiency with Large Language Models (LLMs), prompt engineering at scale, reasoning techniques, context optimization, and response quality tuning is expected.

Solid experience working with agent frameworks and platforms such as LangChain, LangGraph, AutoGen, CrewAI, and/or custom enterprise agent frameworks is crucial.

A deep understanding of agent orchestration and workflow design, including task planning, tool routing, exception handling, fallback, and escalation strategies, is vital.

Experience in designing memory and context management strategies, including short-term/long-term memory, vector stores, stateful execution, and context isolation, is required.

A strong background in Retrieval-Augmented Generation (RAG), embedding pipelines, vector databases, enterprise knowledge grounding, and search optimization is necessary.

Experience with algorithmic and model-level ownership is expected.

Hands-on experience deploying and operating agentic systems on Google Cloud Platform (GCP), leveraging cloud-native AI and scalable deployment architectures, is required.

Practical exposure to DevOps and MLOps for AI systems, including CI/CD pipelines, agent and prompt versioning, testing frameworks, and controlled releases, is essential.

Experience implementing observability and monitoring for AI systems, including logging, tracing, latency analysis, quality evaluation, and cost tracking, is required.

A strong understanding of security, privacy, and governance for AI systems, including Responsible AI frameworks, access control, compliance, and auditability, is crucial.

Preferred experience in the General Insurance domain (Health, Motor, Travel), with the ability to map domain workflows to intelligent agent solutions, would be advantageous.

Soft skills include strong ownership, accountability, adaptability, a solution-oriented mindset, effective communication with diverse stakeholders, collaborative leadership, and professional integrity.