Intelligent Optimization Lead, Continuous Improvement
Microsoft (Eightfold Apply)
- Location
- United States, Multiple Locations, Multiple Locations
- Work model
- On-Site
- Level
- Senior
- Posted
- 2h ago
About this role
Overview
The Frontier Transformation Framework helps our customers become Frontier Firms: organizations where AI is part of the operating system the business runs on. The difference between a transformation and a one-off deployment is what happens after go-live. Without a deliberate improvement and optimization discipline, returns erode: adoption plateaus, agents drift, prompt libraries go stale, and high-value agents are starved while low-value ones persist. Intelligent Optimization is the operating discipline that makes AI value compound over time rather than dissipate. The Frontier Transformation team is how the Microsoft Frontier Company, the organization this role sits within, delivers this: a small, senior human-agent team in which each member carries one of the Framework's eight capabilities into the customer, directs a set of digital co-workers to deliver at a scale headcount alone could not, and is designed to fade into a customer counterpart who can carry the work forward. We are seeking a Intelligent Optimization Lead, Continuous Improvement to make AI value compound across strategic customer engagements. In this role, you will establish and run the Intelligent Optimization Framework as a permanent operating rhythm, grounded in kaizen and strong facilitation, that redesigns and reruns workflows to outcomes. You will solve the last-mile problem, embedding what gets built into how work actually gets done through standard work, clear ownership, and an execution cadence that makes new ways of working stick, working directly with business owners, operators, champions, and the Value, Skilling, Change, and engineering leads. This is a role for people who are energized by making change hold rather than launching it, who can read measurement signals and turn them into the next round of improvement, and who believe a program that ships once and leaves is just a release, not a transformation. You are designed to fade into a customer counterpart who owns the improvement rhythm after the engagement ends. If you are excited by turning AI deployments into value that compounds cycle after cycle, this role is for you.
Responsibilities
Establish and run the Continuous Improvement Framework as a permanent, always-on operating rhythm rather than a one-off activity, grounded in kaizen and strong facilitation. Redesign and rerun workflows to outcomes, running structured improvement cycles in which each measurement cycle feeds and improves the next. Solve the last-mile problem by embedding new ways of working into standard work, clear ownership, and a disciplined execution cadence. Read the Copilot Dashboard and other measurement signals to find where usage is shallow, which roles are stuck, and where behavior change is or is not landing, and convert those signals into prioritized improvements. Continuously refine prompt libraries, enablement materials, and agent configurations based on usage data and Work IQ behavior signals. Detect and correct agent drift, retire or rework low-value agents, and direct further investment toward high-ROI agents. Establish standard work, ownership, and review cadences that make improvements durable and repeatable across teams. Facilitate improvement events and retrospectives that treat friction and feedback as inputs to the next cycle rather than one-off fixes. Capture reusable improvement patterns and playbooks that raise the quality and speed of every subsequent cycle. Transfer the continuous improvement discipline, cadences, and tooling to the customer's own people so the rhythm is sustained after the engagement fades. Recommend and drive the discontinuation of workflows, agents, and enablement assets that are not returning value (including work the engagement itself originally built) to ensure capacity flows to the most important value. Success in this role looks like: AI value compounds cycle after cycle rather than plateauing after initial deployment. Adoption deepens, agents stay tuned and