AI Engineer, Forward Deployed (Copilot and Azure AI Foundry)

Collectiv•Chicago, IL

1d•Remote

About The Position

The AI Engineer is a hands-on builder who works directly with Client teams to design, build, and ship production AI solutions on Microsoft 365 Copilot, Copilot Studio, and Azure AI Foundry (now Microsoft Foundry). This is an individual contributor role. You will write the code, configure the agents, wire up the data, and get solutions into the hands of real users. You will work inside a Collectiv delivery team alongside a project lead, architects, and data engineers. You will partner directly with Client business and IT stakeholders to turn a business problem into a working AI solution, then harden it for production.

Requirements

Bachelor's degree from an accredited university in Computer Science, Data Science, Engineering, or a related field (or equivalent experience)
4+ years of experience in software engineering, data engineering, or Microsoft 365 and Power Platform development
1+ years of hands-on experience building and deploying generative AI or LLM-powered solutions in production or Client environments
Hands-on experience with Microsoft 365 Copilot extensibility and Copilot Studio
Hands-on experience with Azure AI Foundry or Azure OpenAI
Experience building RAG solutions with Azure AI Search
Strong Python or C# skills
Experience working in Azure (storage, compute, identity, networking)
Ability to work independently in a Client-facing consulting role

Nice To Haves

Experience with Microsoft Agent Framework, Semantic Kernel, or AutoGen
Experience with Power Platform, Dataverse, and Power Automate
Experience with Microsoft Fabric, Databricks, or Power BI
Experience with LLMOps practices including evaluation frameworks, prompt management, and model monitoring
Familiarity with security, privacy, and compliance considerations for AI in regulated industries
Consulting experience across multiple Clients or industries
Relevant certifications such as Azure AI Engineer Associate, Copilot Studio or Power Platform certifications, or Fabric (DP-600, DP-700)

Responsibilities

Build declarative agents, Copilot connectors, and plugins that extend Microsoft 365 Copilot with Client data and workflows
Build Copilot Studio agents, topics, actions, and autonomous agents for business processes
Connect Copilot to SharePoint, Teams, Microsoft Graph, Dataverse, and line-of-business systems
Write and iterate on prompts, instructions, and agent configurations based on test results and user feedback
Build custom agents and AI applications in Azure AI Foundry for use cases that go beyond out-of-the-box Copilot
Select, deploy, and configure models from the Foundry model catalog, and explain the trade-offs in cost, quality, and latency
Implement agentic workflows with tool calling, multi-step orchestration, and human-in-the-loop controls
Integrate AI solutions with enterprise systems through APIs, Azure Functions, Power Automate, and Logic Apps
Build end-to-end retrieval-augmented generation (RAG) pipelines including ingestion, chunking, embedding, vector and hybrid search, and grounding
Use Azure AI Search and Azure AI Document Intelligence to index and extract knowledge from Client content
Partner with Collectiv data engineers to connect AI solutions to Microsoft Fabric, Databricks, and other enterprise data platforms
Build evaluation sets and test harnesses that measure answer quality, groundedness, hallucination rate, safety, latency, and cost
Implement logging, monitoring, and tracing for deployed agents and models
Apply security, identity, and governance controls using Microsoft Entra ID, Microsoft Purview, and Azure AI Content Safety
Use version control, CI/CD, and code review standards for all AI solution code and configuration
Work directly with Client stakeholders to gather requirements, run working sessions, demo progress, and manage expectations
Translate business problems into scoped, buildable AI use cases with clear success metrics
Train Client users and administrators on the solutions you build and support adoption after go-live
Contribute reusable patterns, accelerators, and documentation to Collectiv's AI practice