Senior Agentic Workflow Engineer

Overview

Lead architect and builder of HRS's portfolio of AI agents and agentic workflows, defining the multi-agent architecture and development standards across platforms and driving delivery of production-grade agents used by account executives, sales operations managers, and customer success teams.

What you'll do

• Define the platform-agnostic multi-agent topology, conversation and workflow design principles, naming conventions, guardrail patterns, and release standards the team follows across all engines.
• Design the orchestrator and sub-agent structure, handoff protocols, context passing, and fallback behaviors, including cross-platform communication via open standards like MCP and A2A.
• Lead design and delivery of agents for pipeline queries, opportunity stage updates, forecasts, opportunity creation, approval workflows, and AI-generated content across a growing sales and customer-facing backlog.
• Partner with the platform engineer to define tool, API, and connector contracts for Dynamics 365, Dataverse, SharePoint, OneDrive, SAP, Clay, AWS data products, and internal APIs; enforce governed interfaces only, with no direct database access, no free-form SQL, and no plaintext credentials.
• Define and implement RBAC, DLP controls, prompt-injection defenses, output validation, and audit logging standards for every agent on every platform.
• Own versioning, regression testing, and staged rollout to production, including human approval gates for write-back actions, and sign off on every production promotion.
• Mentor the mid-level engineer through architecture advice, reviews, and structured onboarding.
• Work with sales and sales operations to scope use cases, define acceptance criteria, and manage delivery expectations.

What you'll need

• Typically 7+ years in software, automation, or integration engineering, with a track record of shipping AI agents or agentic workflows that real users rely on in production.
• Proven experience deploying production agents, including multi-agent orchestrator/sub-agent architectures, on at least one agentic engine (e.g. Copilot Studio, n8n, Dify, Azure AI Foundry, Amazon Bedrock Agents, LangGraph).
• M-shaped profile: deep agentic engineering expertise plus broad knowledge of enterprise systems such as CRM (e.g. Dynamics 365), ERP (e.g. SAP), Microsoft 365, and cloud data platforms.
• Deep understanding of orchestrator/sub-agent design, prompt engineering, RAG, tool/function calling, MCP, agent memory, and evaluation.
• Clear view of each platform's strengths and limits, designing every agent to make the most of where it runs.
• Confident with REST, JSON, webhooks, and OAuth 2.0 / service principals; consumes APIs, connectors, and MCP servers and specifies precisely what agents need from the platform team.
• Solid knowledge of identity and RBAC (e.g. Entra ID), DLP, secrets management, environment and solution lifecycle, and CI/CD for agents.
• Understanding of GDPR and EU data residency for agent data handling.
• Proven ability to coach engineers and uphold quality standards across a team.
• Strong skills in architecture decision records, conversation and workflow specs, and runbooks.
• Fluent English, verbal and written, able to lead discussions with technical and business stakeholders.

Nice to have

• Production experience on multiple agentic platforms.
• Certifications: Microsoft PL-400 or AI-102, AWS, or platform-specific equivalents.

Details

• Location: Bangalore Office.