Project 04
Keylime Interactive
AI + Upskilling
Experience Studio Project
Keylime Interactive AI Upskilling
Designing AI-Powered Upskilling Tools for Early Professionals and Students
Project Overview
In today's rapidly evolving job market, "upskilling"—the process of learning new, relevant skills—is no longer optional. While Artificial Intelligence presents a massive opportunity to personalize and scale learning, its role is often met with a mix of excitement and skepticism. Our team partnered with Key Lime Interactive (KLI), a woman & minority-owned UX/CX agency, to research and identify opportunities in this complex space.
Team
1 UX Graduate Student, 4 UX Undergraduate Upperclassmen, 4 UX Undergraduate Underclassmen
Project Scope
UX Design
Research
AI
Wireframing
Prototyping
Tools
Figma
FigJam
Timeline
January - May 2025
My Contributions
Foundational Research & Synthesis
Primary Research & User Engagement
Collaborative Ideation & Sketching
Mid-Fidelity Prototyping & Testing
User Testing & Feedback Analysis
Documentation & Reporting
Design Space
Problem Statement
Professionals at all levels know they need to keep learning, but the path is often unclear. Early-career professionals feel overwhelmed by resources, while students struggle to bridge the gap between academic theory and real-world application. Companies invest in learning platforms, but these often fail to address the core human needs for guidance, feedback, and practical experience.
Design Question
How can AI be leveraged to create upskilling solutions that are effective, trusted, and address the nuanced needs of both students and professionals?
Project Goals
Communicate our findings after investigating AI and upskilling, and exploring opportunities into where and how they intersect
Based on our findings, design and iterate a possible solution combining AI and upskilling into mid fidelity
User Group
Upperclassmen Students
Early-Career Professionals
Approach
Our approach was a deep, multi-phase investigation. Starting off with research to build a comprehensive understanding from the ground up. We would take our key research insights into ideation where we would develop sketches and wireframes for concept testing. Our results guided our mid-fidelity prototypes that were used in usability tests. Our final deliverables included our research insights, prototypes, and the results from our usability testing.
Phase One: Primary and Secondary Research
This phase aimed to establish a foundational understanding of AI in the context of upskilling. We conducted secondary research, a comparative analysis of existing platforms, user surveys, and 12 interviews.
Phase Two: Ideation and Sketching
By focusing on hands-on simulators and personalized roadmaps, our features help users build tangible skills and see their progress clearly, directly addressing the desire for real-world application.
Phase Three: Prototyping and Testing
The Feedback Dashboard and Job Simulator provide the clear guidance users are missing, while still allowing them the autonomy to choose which skills to focus on and how to learn them.
Phase One: Primary and Secondary Research
This phase aimed to establish a foundational understanding of AI in the context of upskilling. We conducted secondary research, a comparative analysis of existing platforms, user surveys, and 12 interviews.
Guiding Questions
What are the core user needs and pain points in upskilling?
How do AI tools currently support or hinder progress toward user goals?
Goals
Develop a comprehensive understanding of AI and Upskilling tools
Analyze existing solutions in the market
Identify user motivations and pain points in the upskilling process
Secondary Research
This research sprint aims to explore broader trends, opportunities, and challenges in upskilling and AI. It allows for a comprehensive understanding of the external landscape, which is crucial for strategic planning.
Activities in this sprint include:
Literature reviews
Conducting a comparative analysis
Conducting a SWOT analysis
Literature Review
We split our team into two to look at two major topics: Artificial Intelligence and Upskilling. We looked at both academic articles and user opinions on sites like Reddit to gather a wide swath of data.
Comparative Analysis
We needed insight into the tools that are currently available with a comparative analysis. We researched existing platforms that provide upskilling resources with features that include artificial intelligence.
Specialized Learning
Learning Systems
Skill Development
Career Coaching
SWOT Analysis
Our SWOT is structured around the same four key segments which are strengths, weaknesses, opportunities, and threats as the comparative analysis. This analysis helped us refine further and evaluate the effectiveness of existing AI-driven upskilling solutions, pinpoint gaps, and explore areas for innovation and improvement.
Primary Research
This sprint utilizes both interviews and a survey. The interviews will allow us to collect deep qualitative data about our user groups, and the survey will allow us to collect quantitative data about our most available user group.
Survey
We did not get a representative sample from our target population we would need to make claims on how our user group feels. Instead, we took these patterns and explored them more with secondary research:
Preferred upskilling methods include watching YouTube videos, taking online courses, and more practice through projects.
50% of students who upskill spend 1-2 hours a week and 50% spend 3-5 hours, which shows their willingness to invest their time in upskilling.
69.2% of respondents use AI tools like ChatGPT daily and are open to using AI for learning. However, people still prefer human guidance for career-related learning.
We successfully collected contacts for additional interviews with an incentive. The results of these interviews can be found below.
Interviews
We conducted 12 interviews with people from all user groups to understand the role and perception of AI and key pain points in upskilling. We would interview:
3 upperclassmen
3 early career professionals
3 experienced professionals
3 hiring managers
Key Research Insights
Across all groups, a clear narrative emerged. While AI is seen as a useful tool for efficiency, it is not yet trusted as a teacher.
Human Interaction is Essential: The most valuable learning experiences involve mentorship, peer collaboration, and direct feedback from managers. AI was not seen as a replacement for this.
Low Trust in AI for Critical Tasks: Users were comfortable using AI for basic, repetitive tasks (summarizing, ideating) but were highly skeptical of its ability to provide nuanced feedback, evaluate complex work, or teach unfamiliar topics.
Lack of Structure is a Major Barrier: Both students and professionals struggle with knowing what to learn next. They desire clear, personalized roadmaps but find current AI recommendations too generic.
Soft Skills are Underdeveloped: A significant gap exists in developing crucial soft skills like communication and leadership through digital tools.
Practical Experience > Certifications: Hiring managers and professionals alike value hands-on, real-world project experience far more than certificates from online courses.
Phase Two: Opportunity Exploration and Ideation
Our research revealed that a single "one-size-fits-all" platform would be ineffective. Instead, we identified key opportunity areas and designed a suite of targeted, AI-powered features that could be integrated into a cohesive upskilling ecosystem. The goal was not to replace human interaction, but to augment it.
We would perform extensive ideation and sketching, including a Crazy 8s workshop.
Guiding Question
How can we use our synthesized research to define our opportunities and ideate on potential solutions?
Goals
Define opportunities we identified through both primary and secondary research.
Ideate potential solutions to these opportunities.
Sketch initial low-fidelity solutions that addressed our gaps.
Identifying Gaps
From our research findings, we identified and synthesized primary gaps and opportunities for ideation:
Not enough human interaction and constant feedback
Diverse learning styles (structured vs unstructured)
Building users’ trust in AI (using AI in a more dependable role)
Encourage the learning of soft skills
Efficient knowledge building
Need for experience relevant to real-world projects
Initial Ideation
From our research findings, we identified and synthesized primary gaps and opportunities for ideation:
Crazy 8's Workshop
Each member of the team took one gap that we found earlier in our research synthesis and drew one sketch per minute for 8 minutes. We ended with 64 sketches and ideas to talk about. After sharing, we put our sketches on our FigJam and voted via star stickers on the ideas/sketches that we determined the best.
Through this, we identified three essential features where AI can play a key role in supporting users on their upskilling journeys:
Human Interaction:
Helps users stay connected with peers and build a sense of motivation and accountability throughout their learning process.
Personalization:
Assists users in better understanding their individual needs and clarifies what steps they need to take to make progress in their upskilling journey.
Credibility:
Builds trust by ensuring users feel secure about how their data is handled and confident in the quality and reliability of the resources they use.
Phase Three: User Feedback and Iteration
Taking our low-fi sketches, we looked for feedback through value proposition and usability testing. We iterated our sketches into wireframes and to mid-fidelity prototypes.
Guiding Question
How can we use our previous insights and ideation to create and verify AI in upskilling solutions?
Goals
Gain feedback and insights on sketches through user testing.
Test, refine, and further iterate sketches.
Value Proposition and User Testing
We took our best sketches and tested them with both of our user groups to gather insights into what they prefer and what they would use.
Final Solutions
We chose the highest scores of low-fidelity sketches from our user testing. We iterated upon them and developed them into mid-fidelity prototypes that communicate our solutions:
Design Rationale
Building Trust
`We position AI as a supportive assistant rather than an authoritative judge. It analyzes data, suggests paths, and provides practice, but the critical feedback and decision-making loops still involve humans (managers, mentors, peers).
Fostering Competence
By focusing on hands-on simulators and personalized roadmaps, our features help users build tangible skills and see their progress clearly, directly addressing the desire for real-world application.
Shopper Experience
The Feedback Dashboard and Job Simulator provide the clear guidance users are missing, while still allowing them the autonomy to choose which skills to focus on and how to learn them.
Limitations
Potential for AI Bias: Our participants were largely familiar with AI tools. This may have skewed our findings, as their perceptions might not reflect those of professionals who are less experienced or more skeptical of AI.
Limited Industry Input: As a student-led project, our access to a broad spectrum of industry professionals was limited. Deeper collaboration with more companies would be necessary to fully understand specific corporate upskilling needs.
Next Steps
Refine and Integrate Prototypes:
Develop the proposed features into a single, high-fidelity, and functional prototype to test the cohesive user journey.
Conduct Broader User Testing: Validate the solutions with a larger, more diverse user pool to ensure the features are effective and trusted across different industries and roles.
Explore AI Ethics and Feasibility:
Conduct a deeper investigation into the technical and ethical implications of implementing these AI models, ensuring fairness, transparency, and data privacy.