GHR AI Engineer – Agentic Integrations (MCP / A2A / Python / APIs) AVP

GHR AI Engineer – Agentic Integrations (MCP / A2A / Python / APIs)

Corporate Functions Technology

Reports to: Managing Director, Corporate Functions Technology

Role Overview

The GHR AI Engineer – Agentic Integrations will help build and integrate employee-facing AI capabilities by connecting a custom React intranet experience to internal AI services and to external vendor agents. This is a hands-on engineering role focused on building the “integration layer” for agentic systems: Python services, APIs, connectors, tool definitions, and secure runtime patterns that allow agents to take actions and retrieve knowledge safely.

The ideal candidate has strong Python fundamentals, comfort with APIs and authentication, and practical experience with Model Context Protocol (MCP) and agent-to-agent collaboration patterns. You will partner closely with front-end engineers, solution architects, cybersecurity, and platform teams to deliver secure, scalable AI integrations that improve employee experience (with GHR/HR domain knowledge as a plus).

Key Responsibilities

Agentic Integration Engineering (MCP + A2A)

• Build and maintain MCP-based connectors that expose internal tools and data sources to AI agents in a consistent, secure way (tools/resources/prompts).
• Implement agent-to-agent interaction patterns for multi-step workflows where agents coordinate across internal systems and external vendor agents.
• Define tool contracts (schemas, parameters, outputs), versioning, and compatibility practices for long-term maintainability.

Internal AI Gateway & Vendor Agent Connectivity

• Integrate your intranet and internal platforms with the enterprise AI gateway and approved LLM/agent runtimes.
• Connect to external vendors that provide agents or agent-accessible capabilities (via MCP or API-based tool calling) while enforcing authentication, authorization, and auditing requirements.
• Implement safe, reliable request routing patterns (timeouts, retries, circuit breakers, fallbacks) and error-handling standards.

API & Backend Development (Python)

• Develop lightweight Python services that broker requests between the React intranet and agent runtimes/tools.
• Build and maintain REST APIs (and/or async workers) that support agent workflows, tool execution, and structured responses back to the UI.
• Support data shaping and transformation (JSON, schemas, normalization) to make multi-source responses usable for employee-facing experiences.

Grounding, Retrieval & Knowledge Enablement (as needed)

• Assist in implementing retrieval/grounding patterns (document ingestion, chunking, embeddings, retrieval, citations) for intranet use cases.
• Integrate knowledge sources (policies, HR content, FAQs, service catalogs) while respecting access controls and data classification requirements.

Security, Compliance & Safe-by-Design Engineering

• Implement least-privilege access for tools and connectors, secure secrets management, and strong logging/audit trails for tool invocation.
• Help harden integrations against prompt injection and tool misuse by adding confirmation steps for sensitive actions, validation of tool inputs, and content safety checks.
• Partner with Cybersecurity and Risk teams to ensure solutions meet enterprise standards and are audit-ready.

Observability, Testing & Operations

• Instrument services with logging, metrics, and traces; support monitoring dashboards and alerting.
• Contribute to test automation (unit/integration), regression testing for tool definitions, and release readiness practices.
• Support production triage and continuous improvement after releases.

Collaboration & Communication

• Work closely with the React intranet team to define API contracts and deliver features end-to-end.
• Produce clear documentation (tool specs, runbooks, integration diagrams) and communicate progress/risks effectively.
• Participate in Agile ceremonies and deliver incremental value in short cycles.

Experience & Competencies

• 3–5 years of hands-on software engineering experience, with strong Python development skills.
• Practical experience with APIs (REST), JSON/schema-driven payloads, and common auth patterns (OAuth/OIDC, token handling).
• Familiarity with agentic concepts: tool/function calling, structured outputs, multi-step orchestration, and safe execution patterns.
• Experience with MCP and/or agent-to-agent communication patte…