AI Innovation Analyst

MIG Real Estate is seeking an AI Innovation Analyst to design, build, and operate AI agents and automated workflows that improve the efficiency and accuracy of MIG's business operations, leading these initiatives as discrete projects from discovery through deployment. Reporting to the CTO, you will partner closely with asset management, accounting, portfolio management, and acquisitions teams. This is an early-career, remote role with significant career runway.

About The Job

At MIG Real Estate, this role sits at the center of how a multi-billion dollar, multi-asset institutional investor reinvents its own operating model. You will design and ship the agents and automations that change how MIG's teams underwrite, manage, and report across a multi-asset portfolio. The position offers direct CTO partnership, high autonomy, and a career path within a function still being defined.

This role is part of a broader buildout of MIG's analytics and AI capabilities, complementing the automation work already underway on MIG's Business Intelligence team with dedicated process design and platform expertise.

The ideal candidate is a process-minded, business-curious builder who understands that an agent's value depends less on the underlying model than on how cleanly it integrates with existing systems, data, and workflows. You'll build primarily within MIG's sanctioned platforms, including Claude Code, Claude Cowork, and Microsoft Copilot Studio, rather than developing and hosting standalone applications. Our focus is clean, secure integration and reliable orchestration.

Initial work will center on internal operations — reporting, document handling, data validation, and routine analysis — before progressing to externally-facing or transactional workflows, in close coordination with the Business Intelligence team and business stakeholders.

Responsibilities

• Design, build, test, and deploy AI agents and automated workflows across MIG's business functions, starting with internal operations before progressing to externally-facing or transactional workflows.
• Compose agents on MIG's sanctioned platforms — Microsoft Copilot Studio, Azure AI Foundry, Anthropic's Claude (Cowork, Claude Code, and Managed Agents), and/or OpenAI's Custom GPTs and the Responses API with the Agents SDK.
• Develop and maintain prompts, tool definitions, and orchestration logic that make agents reliable and predictable in production.
• Establish guardrails, evaluation pipelines, and monitoring so agent behavior is measurable and observable, not just demo-ready.
• Partner with business teams to identify high-value automation opportunities; conduct structured discovery sessions to map current-state workflows, surface bottlenecks, and assess data and process maturity.
• Translate business problems into agent specifications: required inputs, allowed actions, decision logic, escalation paths, and success criteria.
• Document the happy path, failure modes, and human-in-the-loop checkpoints for every deployed agent.
• Work closely with the Business Intelligence team to connect agents to MIG's data infrastructure: Microsoft Fabric, OneLake, ADLS, SharePoint, Power BI semantic models, Yardi Voyager, RealPage, and other systems.
• Develop and maintain integrations using REST APIs, Power Automate flows, Azure Logic Apps, and MCP servers, with rigor around data lineage, permissions, and read/write boundaries.
• Lead agent and automation initiatives as discrete projects: define scope and success criteria, sequence work, manage timelines and dependencies, coordinate stakeholders across IT and business teams, and maintain a prioritized backlog of opportunities.
• Provide regular status reporting and demos to the CTO and business stakeholders, and drive adoption through end-user training, feedback, and iteration.
• Adhere to MIG's AI policy, data protection standards, and SSO requirements; maintain documentation of every deployed agent (purpose, owners, data accessed, prompts, evaluation results, security review and identified risks, known limitations) and contribute to MIG's evolving Data Governance framework.

Requirements

• Disciplined about scope, requirements, and deliverables — does not start building before understanding the problem; comfortable mapping a workflow, identifying weak points, and proposing a sequenced solution.
• Naturally documents decisions and design choices for the next person; writes clearly, and treats documentation as part of the job rather than an afterthought.
• Demonstrated project management instincts: comfortable scoping work, sequencing tasks, managing stakeholders, and keeping a portfolio of concurrent initiatives moving without dropping the ball. Formal PM certification not required.
• Curious about how a business actually operates; able to sit with an asset manager or controller, listen carefully, and understand their real problem.
• Hands-on experience building automations with at least one modern platform (Power Automate, Copilot Studio, or similar).
• Comfortable with Python or another scripting language sufficient to wire APIs together, manipulate JSON, and write small utility scripts.
• Basic understanding of REST APIs, authentication patterns, and webhooks.
• Familiarity with prompt design, LLM tool/function calling, and at least one agent framework or orchestration platform.
• Working knowledge of SQL and data manipulation; ability to read a data model and understand what a field actually represents in production.
• Able to explain a technical concept to non-technical colleagues.

Qualifications

• 1–3 years of relevant experience, including within the Microsoft 365 ecosystem (SharePoint, Teams, Power Platform, Azure).
• Bachelor's degree in a quantitative, business, or technical field, or equivalent demonstrated experience.
• Demonstrated ability to ship working software or automations end-to-end — portfolio, GitHub, or specific deployed examples preferred over coursework.
• Comfort working in a remote-first environment.
• Advantages but not required: (1) Prior exposure to real estate or financial services, (2) Experience integrating with Yardi, MRI, RealPage, or comparable industry systems and/or (3) Familiarity with MCP (Model Context Protocol) and agent frameworks (LangChain, LangGraph, CrewAI, Semantic Kernel).