MCP and Enterprise Integration Engineer

Job description

NTT DATA strives to hire exceptional, innovative and passionate individuals who want to grow with us. If you want to be part of an inclusive, adaptable, and forward-thinking organization, apply now. We are currently seeking a MCP and Enterprise Integration Engineer to join our team in Hyderabad, Telangana (IN-TG), India (IN). - MCP and Enterprise Integration Engineer **Model Context Protocol | Enterprise tools and APIs | Secure, reusable integration** **Number of positions** 1 **Level** Senior Individual Contributor **Primary locations** Hyderabad or Noida preferred; exceptional onshore candidates may be considered **Target / alternate titles** MCP Engineer; AI Integration Engineer; GenAI Integration Engineer; Enterprise API Engineer - AI; Tooling Engineer - Agentic AI; Backend Integration Engineer **Core keywords** Model Context Protocol, MCP server, MCP tools, tool schemas, JSON Schema, APIs, microservices, enterprise integration, OAuth2, OIDC, mTLS, service accounts, identity, authorization, secrets, Java, Python, TypeScript, ServiceNow, SailPoint, CMDB, event integration, observability **Recruiter red flags** Only consumed MCP tools; no server implementation; generic middleware profile with no agent or LLM understanding; weak identity and authorization knowledge; no production API ownership; no contract testing or observability. **Role purpose** Design, build, secure, and operate MCP servers, tools, and enterprise-integration components that allow agents to interact safely with business systems, data platforms, knowledge sources, and operational services. The role will establish reusable tool contracts and integration patterns that can be extended by internal teams as new ecosystems and use cases are added. **Client and delivery context** - The first MCP server and tools create the pattern for broader scale. Once the scaffolding is proven, internal teams should be able to add tools with lower effort and risk. - The client already applies strong review and security scrutiny to production MCP services. The engineer must produce implementation and evidence that can pass established controls. - Integrations may span service management, identity, configuration data, knowledge bases, data platforms, business applications, and custom internal APIs. - The engineer must understand both agent behavior and traditional enterprise integration; tool exposure must not bypass business rules, authorization, or auditability. **Primary Ownership** - MCP server architecture and implementation, tool and resource definitions, schemas, discovery, versioning, lifecycle management, and developer documentation. - Secure connectivity to enterprise systems, including authentication, authorization, secrets, service identities, network controls, and data access enforcement. - Reliable adapters, APIs, events, and integration services that expose controlled capabilities to agents. - Tool-level observability, testing, error management, performance, resilience, and reusable implementation patterns. **Key Responsibilities** - Design and implement MCP servers, tools, resources, prompts, schemas, and capability-discovery patterns aligned with enterprise standards. - Translate enterprise API and workflow capabilities into narrowly scoped, well-described, validated, and auditable agent tools. - Implement OAuth2/OIDC, mTLS, service identities, delegated authorization, role or attribute-based access, secrets management, and least-privilege controls. - Build connectors and adapters for enterprise platforms, knowledge systems, databases, event streams, file repositories, and custom services. - Implement input validation, output normalization, idempotency, retries, timeouts, rate limits, circuit breakers, pagination, caching, and consistent error contracts. - Prevent unsafe tool use through explicit schemas, server-side business rules, authorization checks, confirmation patterns, and restricted action surfaces. - Create contract, integration, security, performance, and end-to-end tests for MCP tools and downstream dependencies. - Instrument tool calls with tracing, metrics, structured logs, audit events, latency, failure reasons, and dependency health. - Support deployment, release management, version compatibility, rollback, incident troubleshooting, and production support for MCP services. - Publish reusable SDKs, templates, tool-development guidance, examples, and onboarding documentation for internal teams. **Must-have candidate profile** - 7+ years of backend, API, middleware, platform, or enterprise-integration engineering experience. - Hands-on experience building MCP servers or equivalent agent tool and function-integration layers; direct MCP experience is strongly preferred. - Strong development skills in Python, Java, TypeScript/Node.js, Go, or a comparable backend language. - Deep understanding of API design, JSON Schema, contracts, microservices, distributed systems, asynchronous integration, and error handling. - Strong identity and security experience, including OAuth2, OIDC, tokens, service accounts, mTLS, authorization, secrets, and audit logging. - Experience integrating with enterprise applications, data systems, knowledge platforms, or operational tooling in production. - Experience with automated testing, API gateways, containers, CI/CD, logging, tracing, monitoring, and production troubleshooting. - Understanding of LLM tool calling, agent risks, prompt injection, data leakage, excessive agency, and server-side control requirements. **Preferred Experience** - Experience integrating with platforms such as ServiceNow, SailPoint, CMDB systems, collaboration suites, document repositories, search platforms, or data catalogs. - Experience with API management, service mesh, event streaming, queues, webhooks, and enterprise integration platforms. - Experience developing reusable SDKs, connector frameworks, internal developer platforms, or integration