AES - DE - Generative AI Application Developers

QUICK FACTS

EngagementZensar at Client (Client Site)LocationHybrid / On-site - Client Engineering HubsSeniorityMid to Senior (4-8 years)EmploymentFull-Time Contract with conversion pathGrowth PathMCP Build Internal AI Platform Engineering

KEY RESPONSIBILITIES

MCP Server Design & Development

• Design and implement MCP (Model Context Protocol) servers in Python, exposing enterprise tools and internal APIs as Claude-accessible resources and tool calls

• Build MCP integrations for Client's existing internal stack - Jira, GitHub, Confluence, Salesforce, internal data APIs, and custom microservices

• Implement both SSE (HTTP/streaming) and stdio transport modes depending on deployment context, and advise teams on when to use each

• Design robust tool schemas - well-defined input/output contracts, clear tool descriptions that guide Claude's reasoning and usage

• Write test suites for MCP servers - unit tests, integration tests with MCP Inspector, and end-to-end validation with Claude Desktop

Authentication, Security & API Integration

• Implement OAuth 2.0 flows (Authorization Code, Client Credentials, PKCE) for secure MCP server authentication - following the MCP authorization spec

• Integrate with identity providers (Okta, Azure AD, Google) to enable SSO-based access control on MCP servers

• Design and implement API gateway patterns for MCP backends - rate limiting, scoped token management, audit logging

• Ensure MCP servers meet enterprise security standards - secrets management (Vault, AWS Secrets Manager), TLS, least-privilege access

• Build adapters for REST, GraphQL, and gRPC-based internal APIs, abstracting complexity behind clean MCP tool interfaces

Platform Engineering (Growth Path)

• Contribute to the design of Client's internal AI platform - a shared infrastructure layer for deploying, discovering, and managing MCP servers at scale

• Build developer-facing tooling: CLI utilities, SDK wrappers, scaffolding templates that make it fast for Client engineering teams to build new MCP integrations

• Implement observability for the MCP layer - structured logging, distributed tracing, dashboards (Datadog, Grafana) to monitor AI tool usage across teams

• Design multi-tenant MCP deployment patterns - namespace isolation, per-team credential scoping, usage quotas

• Work with Client's platform team to containerize and deploy MCP servers on Kubernetes, with CI/CD pipelines and GitOps workflows

Collaboration & Enablement

• Act as the technical MCP subject-matter expert for Client's engineering teams - running office hours, reviewing integration designs, unblocking builders

• Collaborate with Endpoint AI Support Engineers (Role ZEN-RBK-ENG-01) to ensure seamless end-to-end experience from user machine to MCP server

• Write technical documentation, integration guides, and architecture decision records (ADRs) for all MCP infrastructure

• Participate in Client's AI working group - contributing insights from the integration layer to shape overall AI strategy

REQUIRED SKILLS & EXPERIENCE

Backend Engineering

• 4+ years of Python backend development - FastAPI, Flask, or similar async frameworks; clean, testable, production-grade code

• Strong REST API design skills - resource modeling, HTTP semantics, versioning, pagination, error standards (RFC 7807)

• Experience consuming and building integrations with third-party APIs (SaaS platforms, internal microservices)

• Proficiency with async Python (asyncio, httpx) - critical for MCP server performance

• Node.js/TypeScript familiarity is a strong plus - the MCP SDK has first-class TypeScript support

Authentication & Security

• Deep understanding of OAuth 2.0 - grant types, token introspection, refresh flows, scopes

• Experience integrating with OAuth/OIDC identity providers in production: Okta, Azure AD, or Google Workspace

• JWT handling - signing, validation, claims inspection, expiry management

• Secure secrets management - environment variables, secrets vaults, never hardcoded credentials

Infrastructure & DevOps

• Containerization with Docker - writing production Dockerfiles, multi-stage builds, image optimization

• Kubernetes basics - Deployments, Services, ConfigMaps, Secrets, Ingress; comfortable reading and writing YAML manifests

• CI/CD experience - GitHub Actions, GitLab CI, or similar; automated testing and deployment pipelines

• Cloud-native mindset - AWS, GCP, or Azure; familiarity with managed services (Lambda, Cloud Run, ECS)

AI & MCP Ecosystem

• Working knowledge of MCP (Model Context Protocol) - understanding of the protocol primitives: tools, resources, prompts, sampling

• Experience with the Anthropic Python SDK or Claude API - making API calls, handling streaming responses, function calling/tool use

• Awareness of LLM integration patterns - prompt engineering basics, context management, tool result handling

• Familiarity with agent frameworks (LangChain, LlamaIndex, or similar) is a plus

NICE TO HAVE

• Prior experience building MCP servers - even personal/open-source projects are highly valued

• Contributions to open-source MCP server repositories or the MCP spec discussion

• Background in developer tooling, internal platforms, or API gateway products

• Experience at a SaaS or security company (highly relevant given Client's domain)

• GraphQL API design and federation

• Familiarity with Anthropic's Claude system prompt design and tool-use best practices

SKILLS AT A GLANCE

Python

FastAPI

OAuth 2.0

MCP Protocol

REST APIs

Docker / K8s

Claude API

Platform Eng