Agentic AI Engineer

Overview

Agentic AI Engineer is a hands-on engineering role in the Data and AI Office, reporting to the Head of AI Engineering. The role builds and operationalises secure, production-grade AI agents, Copilot extensions and reusable enterprise capabilities across the delivery lifecycle.

What you'll do

• Work with product owners, business SMEs, data owners and risk stakeholders to translate business processes and user needs into technically viable agentic AI solutions with clear outcomes, acceptance criteria and operational constraints.
• Contribute to the technical design of assigned agents and Copilot solutions, including secure grounding and retrieval, identity, access, reliability, maintainability, observability, rollback and supportability.
• Design and write production-quality code, configurations, prompts, agent instructions, tools, APIs, connectors, MCP services and automation using Microsoft Foundry, Copilot Studio, Microsoft 365 Copilot extensibility, Microsoft Graph and Copilot APIs, Azure integration services and Power Platform.
• Build reusable components, skills, plugins, tool definitions, prompt templates, evaluation assets, connectors and deployment modules, integrating with governed enterprise data and semantic layers.
• Execute unit, component, integration, regression, grounding, adversarial, performance, accessibility, security and compliance tests.
• Build representative evaluation datasets and business-specific quality rubrics.
• Deploy solutions through controlled development, test and production environments.
• Configure telemetry, traces, dashboards, alerts, auditability and production runbooks.
• Support diagnosis and remediation of production incidents.
• Apply source control, peer review, version control, dependency management, automated build and CI/CD practices.
• Contribute production learning to reusable standards and engineering guidance.

What you'll need

• Agentic AI architecture, orchestration and multi-step workflow design.
• Microsoft Foundry Agent Service and SDK-based development.
• Microsoft Copilot Studio, Power Platform environments, solutions and ALM.
• Claude Code and Codex.
• Microsoft 365 Copilot extensibility, including agents, skills, plugins, connectors, MCP and APIs.
• Python and TypeScript/.NET engineering for production services and integrations.
• API, event, workflow and integration architecture, including Azure API Management, Functions and Logic Apps where appropriate.
• Enterprise grounding, RAG, vector search, semantic models and governed data products.
• Microsoft Entra identity, OAuth, managed identities, RBAC and least privilege.
• GitHub or Azure DevOps, CI/CD or configuration automation and release engineering.
• AI evaluation, prompt and tool testing, observability, OpenTelemetry and Application Insights.
• Threat modelling, secure development, privacy, Microsoft Purview and Responsible AI.

Nice to have

• Experience developing and delivering enterprise AI, data, cloud or automation solutions, with evidence of hands-on software engineering.
• Experience contributing across requirements, design, build, testing, go-live and live operations, including integration with enterprise data, APIs, identity and security controls.
• Experience operating in a regulated, privacy-sensitive or similarly controlled environment is desirable.
• Degree in computer science, software engineering or a related discipline, or equivalent practical experience.
• Relevant Microsoft AI, Power Platform, Azure or security credentials are desirable where aligned to the candidate’s specialism; equivalent practical experience is acceptable.

Details

• Location: Bengaluru Campus 31.