Microsoft AI Skills Navigator
Technical and business professionals are under pressure to build AI skills fast, but finding the right learning content is time-consuming and uncertain. AI Skills Navigator addresses this problem by surfacing relevant content and building personalized, adaptive learning plans.
As UX lead, I helped drive product strategy across two phases: from designing a north star vision through to partnering with engineering and business stakeholders to ship the experience.
Company
Microsoft
Timeline
March 2025 - November 2025
Role
UX Lead
Team
Zachary Porter, Matt Kissick, Michi Broman, Krishna Vadrevu, Brittany Mederos, Jessi Magraw, Matan Lehrer
Skills
Product Design, Visual Design, Interactive Prototyping, Stakeholder Management, User Research & Testing, Content Strategy
Discovery
We conducted foundational interviews with stakeholders and users to better understand project goals and key pain points:
Learning online can be isolating. Users should feel supported and engaged when using the platform, regardless of where their learning content is coming from.
Users learning the same skills have varying goals and available time. Learning plans should be adaptable, yet lead to the same tangible outcomes that will be recognized in the workplace.
Microsoft has a large catalog of learning content dedicated to "upskilling" for work. We should surface this content to users in a personalized way, instead of making them hunt for it.
Competitive Analysis
In parallel with the interviews, we conducted a competitive analysis to identify inspirational experiences and uncover gaps in the online education space. These activities revealed key themes that informed our guiding design pillars: Assistive, Personalized, Engaging, Motivating, and Insightful.
Concept Evaluation
Using insights from the discovery phase, we created concept sketches addressing key user pain points to guide feature prioritization. The concepts were centered around the following questions:
How might we provide options tailored to a learner’s role, responsibilities, and experience level?
How might we use AI to deliver content that actively engages learners through exercises and hands-on activities?
How might we enable learners to track their progress in a non-linear way?
Wireframes & Content Strategy
After aligning on which concepts to move forward with during a strategy workshop, we began developing the content strategy and wireframes. The selected concepts focused on the parts of the journey where users experienced the most friction: discovery and progress tracking. A key part of this effort was surfacing the right content at the right time. We partnered with a content strategist and used the LinkedIn Learning (a Microsoft partner) “skills” framework to map user goals to tangible learning outcomes.
Concept Testing Results
01
Learners questioned content accuracy, trusting subject matter experts over aggregated internet data.
02
Adaptive learning felt unfamiliar, so learners needed clearer upfront value; the conversational entry point was initially confusing.
03
Once understood and trusted, learners were excited by the platform’s personalization and real-world skill potential.
Final Designs
After completing concept testing, we adjusted the screens based on feedback and moved into the synthesis phase, where we aligned on a visual design language and built the final screens.
Visual Design
Microsoft wanted to use this project as an opportunity to refresh existing design components in their design system, Fluent. We decided on warm purples, rounded edges, and gradients to make the platform feel approachable and differentiate from the corporate cool tones often found in the space.
Agentic Front Door
A conversational entry point that guides learners in building a customized learning plan. A series of prompts and questions helps learners clarify how they want to learn, build trust, and map their input to personalized content recommendations.
Customized Learning Paths
Because this approach to building a learning plan was novel, it was important to introduce a “path” structure that aligned with learners’ mental models. This helps them anticipate what’s next and understand how individual lessons contribute to their broader goals. We presented content this way to make the experience more engaging and convey flexibility.
Adaptive Learning Content
Adaptive content was identified early on by both stakeholders and users as a key differentiator. We incorporated familiar formats like quizzes and flashcards, along with role-playing exercises for learners looking to practice real-world application.
Skill Hub
A place for Learners to track their learning progress and stay up to date with new learning opportunities. It was important for this page to support nonlinear learning and tracking real skills recognized by the workforce. If learners are struggling with a specific skill, the AI agent can intervene and provide more ways for the user to learn.
Impact and Takeaways
The MVP version of the AI Skills Navigator launched thousands of Learners in November 2025, and set the foundation for AI learning at Microsoft. If given more time, there are multiple things I would have loved to explore further.
Visual Language Alignment
During the build phase, we discovered that the visual design direction would be challenging due to Microsoft’s multiple design systems and backend constraints. As a result, the MVP shipped with some interim inconsistencies, but it would benefit from a more cohesive design language—one that is both innovative and scalable within existing systems.
Content Strategy & Building for AI
The conversational experience encountered challenges when we discovered that not all content was properly tagged. To address this, we introduced a more traditional catalog page for learners who may not find what they’re looking for through the agentic experience. With more time, we would have further partnered with the content strategist to refine the schema across content owners.