Role details

- Number of positions: 1
- Level: Principal Engineer / Hands-on Technical Lead
- Primary locations: Hyderabad or Noida preferred; exceptional onshore candidates may be considered
- Target / alternate titles: Principal AI Engineer; Staff Agentic AI Engineer; Lead GenAI Engineer; Principal LLM Engineer; Hands-on AI Solution Lead; Lead AI Application Engineer
- Core keywords: agentic AI, LLM, GenAI, business-process decomposition, agent patterns, MCP, tool calling, context engineering, RAG, workflow orchestration, Python, Java, TypeScript, APIs, microservices, cloud, Kubernetes, CI/CD, evaluation, observability

Job Summary Role purpose: Lead a compact Agentic AI SWAT team that evaluates business processes, selects the correct agentic or non-agentic solution pattern, and delivers production-grade implementations within an established enterprise AI ecosystem. This is a hands-on technical leadership role: the individual is expected to design, code, review, troubleshoot, and deploy alongside the engineering team rather than operate only as an architect or advisor.

Client and delivery context The client already has enterprise AI infrastructure, production MCP capabilities, security and governance controls, and delivery pipelines. The lead must plug into those capabilities and accelerate execution.

The pod may work across multiple business domains rather than one fixed function. The lead must rapidly understand new processes and guide two or more delivery tracks when needed.

The role must balance rapid implementation with reusable engineering patterns, operational reliability, enterprise controls, and knowledge transfer to internal teams.

The lead should be language-, cloud-, model-, and framework-agnostic and able to work with the client's existing technology choices.

Primary ownership

Process-to-solution assessment, including whether a requirement calls for an agent, deterministic workflow, retrieval, conventional application logic, or a hybrid pattern.

Technical direction for agents, MCP tools, context and memory, enterprise integrations, orchestration, evaluation, observability, and production readiness.

Hands-on contribution to critical code paths, design spikes, integration patterns, code reviews, debugging, and release readiness.

Reusable reference implementations,



engineering standards, and knowledge-transfer assets that internal teams can replicate.

Responsibilities

- Work with business and engineering stakeholders to decompose end-to-end processes, identify decision points, data dependencies, controls, exceptions, and measurable outcomes.
- Determine where agentic AI is appropriate and define the operating pattern, autonomy boundary, tool set, context strategy, human oversight, and fallback behavior.
- Design and build production agents and supporting services, including APIs, microservices, event handlers, retrieval components, workflow logic, user-facing interfaces, and integration adapters.
- Define and review MCP server and tool patterns, tool schemas, authentication, authorization, error handling, versioning, and observability.
- Guide multi-agent and workflow orchestration, model selection or routing, state management, retries, timeouts, idempotency, queues, and exception management.
- Establish structured evaluation for task success, groundedness, tool correctness, safety, latency, cost, user acceptance, and regression prevention.
- Contribute directly to code, conduct code and design reviews, resolve complex technical issues, and help engineers move features through test and production environments.
- Align implementations with existing security, architecture, data, privacy, Responsible AI, and production-approval processes without duplicating established governance functions.
- Mentor senior engineers, split work across parallel use cases, manage technical dependencies, and maintain a high bar for engineering quality and delivery pace.
- Create reusable patterns, runbooks, implementation guidance, and technical documentation for subsequent internal scaling.

Must-have candidate profile
- 10+ years of software, platform, AI/ML, or distributed-systems engineering experience, including recent hands-on coding responsibility.
- Demonstrated production delivery of LLM,



RAG, agentic AI, AI assistant, workflow automation, or intelligent application solutions.
- Strong proficiency in at least one enterprise programming language such as Python, Java, or TypeScript, with the ability to work across the broader stack as required.
- Deep understanding of agent patterns, tool calling, context engineering, state and memory, RAG, structured outputs, human-in-the-loop controls, and failure-mode design.
- Experience designing APIs, microservices, event-driven systems, data integrations, authentication patterns, and cloud-native applications.
- Practical experience with CI/CD, containers, Kubernetes or equivalent runtimes, automated testing, logging, tracing, observability, and production support.
- Ability to explain technical trade-offs across quality, latency, cost, security, portability, reliability, and implementation speed.
- Robust communication skills with the credibility to work directly with highly technical client stakeholders and senior engineers.

Preferred experience
- Experience establishing or leading an Agentic AI pod, innovation factory, engineering SWAT team, or accelerated delivery team.
- Hands-on experience developing MCP servers and tools or equivalent standardized enterprise tool-integration layers.
- Experience with model gateways, multi-model routing, policy-based model selection, private models, or provider abstraction.
- Experience in financial services, data and analytics platforms, regulated enterprises, or environments handling sensitive proprietary data.
- Experience across two or more public-cloud ecosystems and multiple commercial or open-source model providers.
- Experience transferring reusable patterns to internal engineering teams and scaling from initial use cases to a broader program.

Indicative technology exposure Python, Java, TypeScript/Node.js; FastAPI, Spring Boot, or equivalent services; LangGraph, Semantic Kernel, AutoGen, CrewAI, LlamaIndex, LangChain, or equivalent agent frameworks; MCP SDKs; REST/gRPC/events; vector and enterprise search; relational and NoSQL databases; Kubernetes, containers, serverless; CI/CD; OpenTelemetry and AI tracing; commercial and open-source LLMs. Specific tools are illustrative, not mandatory.

📌 Principal Agentic AI Engineer / Hands-on Technical Lead (Hyderabad)
🏢 NTT DATA Business Solutions
📍 Hyderabad