AI Engineer

Summary

ICONMA is recruiting for a Communications company seeking an AI Engineer. The role focuses on designing, deploying, monitoring, and iterating production-grade AI agents across Microsoft's AI platform for digital commerce and transformation programs. The engineer will collaborate with cross-functional teams to deliver governed, evaluated, and observable agent-based solutions.

Responsibilities

• Design, build and run the agents that automate work across our digital commerce and transformation programs
• The work is built primarily on Microsoft's AI platform, spanning Azure AI Foundry, Azure Open AI, Microsoft Copilot Studio and the Microsoft 365 and Teams surfaces where our teams already work
• This is a production engineering role rather than a prototyping one
• Agents build are expected to clear the client Agentic Framework bar before they reach users: registered in the Agent Registry, authenticated through Microsoft Entra, authorized under least-privilege RBAC, instrumented for observability, and certified through evaluation testing
• Own that path end to end, from the first system prompt and tool schema through deployment, monitoring and iteration
• Work alongside program managers, data engineers, BI developers, solution architects, and security partners, and you will spend as much time defining what good looks like as writing code

Skills

• Microsoft certification such as Azure AI Engineer Associate (AI-102) or an Azure architecture credential
• Experience with Power Platform, including Power Automate and Data verse, or with Microsoft Fabric and Power BI
• Telecommunications, order management, billing or business-to-business commerce domain experience
• Contributions to agent platform tooling, registries, evaluation frameworks or internal developer experience
• The role is hands-on and product-minded: you take a business workflow, decide where an agent genuinely helps, and deliver a governed agent that people use every day
• Strong candidates are comfortable with ambiguity, opinionated about evaluation, and able to explain in plain language what an agent can do, what it cannot, and how they know
• Bachelor's degree in computer science, software engineering, data science or a related technical field, or equivalent hands-on experience
• Four or more years of software engineering experience, including at least one year building LLM-based or agent-based applications that reached production users
• Strong Python skills, with C# or TypeScript an advantage, and fluency with asynchronous APIs, REST and GraphQL, JSON schemas and Git-based workflows
• Hands-on experience with Microsoft AI services, including Azure AI Foundry or Azure Open AI, Azure AI Search, and at least one of Copilot Studio, Semantic Kernel or the Microsoft Agent Framework
• Practical retrieval-augmented generation experience covering chunking strategy, embeddings, vector search, reranking and retrieval evaluation
• Working knowledge of Model Context Protocol or comparable tool-calling and function-calling patterns, including schema design and versioning
• Demonstrated skill in prompt and context engineering, including system prompts, few-shot design and model-specific behavior
• Experience evaluating LLM systems with offline and online methods, regression suites and human review loops, and a clear point of view on what is worth measuring
• Familiarity with Microsoft Entra ID, OAuth2 and JWT, role-based access control and API gateway patterns
• Experience with CI/CD, containers, infrastructure as code and observability tooling such as Azure Monitor, Application Insights and Open Telemetry
• Clear written communication, including the ability to explain agent behavior, limits and failure modes to partners who are not engineers
• Sound judgment about where an agent is the right answer, and the willingness to say plainly when it is not

Qualifications

Must Haves

• Microsoft certification such as Azure AI Engineer Associate (AI-102) or an Azure architecture credential
• Experience with Power Platform, including Power Automate and Data verse, or with Microsoft Fabric and Power BI
• Telecommunications, order management, billing or business-to-business commerce domain experience
• Contributions to agent platform tooling, registries, evaluation frameworks or internal developer experience
• The role is hands-on and product-minded: you take a business workflow, decide where an agent genuinely helps, and deliver a governed agent that people use every day
• Strong candidates are comfortable with ambiguity, opinionated about evaluation, and able to explain in plain language what an agent can do, what it cannot, and how they know
• Bachelor's degree in computer science, software engineering, data science or a related technical field, or equivalent hands-on experience
• Four or more years of software engineering experience, including at least one year building LLM-based or agent-based applications that reached production users
• Strong Python skills, with C# or TypeScript an advantage, and fluency with asynchronous APIs, REST and GraphQL, JSON schemas and Git-based workflows
• Hands-on experience with Microsoft AI services, including Azure AI Foundry or Azure Open AI, Azure AI Search, and at least one of Copilot Studio, Semantic Kernel or the Microsoft Agent Framework
• Practical retrieval-augmented generation experience covering chunking strategy, embeddings, vector search, reranking and retrieval evaluation
• Working knowledge of Model Context Protocol or comparable tool-calling and function-calling patterns, including schema design and versioning
• Demonstrated skill in prompt and context engineering, including system prompts, few-shot design and model-specific behavior
• Experience evaluating LLM systems with offline and online methods, regression suites and human review loops, and a clear point of view on what is worth measuring
• Familiarity with Microsoft Entra ID, OAuth2 and JWT, role-based access control and API gateway patterns
• Experience with CI/CD, containers, infrastructure as code and observability tooling such as Azure Monitor, Application Insights and Open Telemetry
• Clear written communication, including the ability to explain agent behavior, limits and failure modes to partners who are not engineers
• Sound judgment about where an agent is the right answer, and the willingness to say plainly when it is not

Benefits

• Health Benefits
• Referral Program
• Excellent growth and advancement opportunities