Key Lime Interactive
Upskilling Opportunities with AI
Explored how AI can support professional upskilling through research-driven, mid-fidelity design concepts tailored to students and early career professionals
Role
UX Researcher
Team
Jun Hyok Lim
Kaylee Young
Aishwarya Gorantla
Mahi Tripathi
Mustafa Arshad
Tuan Nguyen
Ashton Sun
Arya Qiu
Duong Le
Tools
Figma, Adobe
Date
Jan - May 2025
Project Description
How might we use AI to help close skills gaps for students seeking career opportunities and early career professionals?
AI is becoming an increasingly significant part of how people learn new things, develop skills, and enhance their existing knowledge. Working with Key Lime Interactive, this project examined how AI can support upskilling and help students and early-career professionals build the skills they need.
We studied how people learn new skills, the challenges they face, and how they currently utilize AI in this process. We also explored how much people trust AI and where they feel comfortable using it, uncovering ways to design solutions that close skill gaps while staying transparent and user-friendly.
Problem
Students and early-career professionals need clearer, more personalized, and trustworthy paths to upskilling
Students and early-career professionals often struggle to identify their skills gaps, find structured and personalized learning paths, and place trust in AI tools that feel disconnected from real-world needs.
Research
Identify board opportunities, trends, and challenges in upskilling and the use of AI
Reviewed academic articles and user discussions (Reddit) to learn users' broad perspectives. The goal was to explore current applications of AI in education, understand how people learn and the challenges they face, and identify how AI can enhance the upskilling process.
Key Insights
Learning method
Most effective when it involves hands-on practice, mentorship, and peers collaboration.
Sucessful factor
Depends on identifying skill gaps, tailoring learning content, and tracking progress.
Barriers
time constraints, limited resources, high costs, and lack of organizational support.
Strengths
strong at automation, insight generation, and engagement
Current uses
In education, it is used for personalized learning paths, explanations, and feedback
Challenges
Unclear context, inaccurate outputs, privacy concerns, and transparency.
Comparative Analysis
How are different platforms using AI to support learning and training?
Conducted a comparative analysis of multiple platforms to understand their strengths, weaknesses, opportunities, and the role of AI in each.
Findings
Cornerstone
Highly customizable with certifications and course options.
Complex, costly, and hard to use for admins and learners.
Docebo
Flexible platform usable across industries.
Customization and features less extensive than advertised.
PluralSight
Large library of beginner-friendly tech and certification courses.
Inconsistent quality and limited depth for advanced learners.
Taskade Skill
Quickly identifies skill gaps with customizable reports.
Expensive and requires learning curve to use effectively.
Career Copilot
Offers tailored career guidance with resume and interview support.
Privacy and data accuracy concerns.
Coach by Coursera
Improves course outcomes and is affordable via subscription.
Lacks differentiation and discoverability of top content.
Edthena
Supports teachers with AI-driven, secure reflection tools.
Limited to analyzing current practice, not teaching new skills.
Fluently
Provides real-time, personalized language feedback during calls.
Accuracy risks as a new product with potential AI errors.
Key Insights
AI is helpful but limited
It’s suitable for support tasks like feedback and recommendations, but not trusted for deep learning.
Human support
People prefer guidance from mentors, coaches, and peers.
Cost and accessiblity matter
Some tools are expensive and mainly for organizations, while cheaper ones lack depth.
Trust is AI is weak
Concerns about accuracy, privacy, and generic advice are common.
User interview
How do students and early-career professionals actually learn new skills, and where do they struggle?
To understand how students and early-career professionals approach upskilling, the challenge they face, and their perceptions of AI as a tool in the learning process
Conducted 12 semi-structured interviews with students and young professionals.
Explored their experiences with learning new skills (academic, professional, or personal).
Asked about pain points, motivations, and how they currently use AI tools.
Key Insights
Hands-on First
Learners prefer projects, practice, and mentorship.
Unclear Path
Unsure what skills to learn or how to track progress.
Trust Gap
Accuracy and context issues reduce confidence in AI.
Time Constraints
Busy schedules make upskilling difficult.
AI as Helper
Used for quick answers, not deep learning.
Human Support
Feedback from peers/mentors keeps learners motivated.
“The hardest part is applying what I learn in a course to real-world work. It doesn’t always stick.”It
“I rely on AI tools like ChatGPT to get started, especially for brainstorming or coding shortcuts.”
“AI helps speed things up, but I can’t fully trust it. It skips details and sometimes gets things wrong.”
Ideation & Sketching
How might we design solutions that make upskilling more practical, personalized, and accessible?
After gathering valuable insights from our research, we moved into the ideation phase. Here, our goal was to translate key findings into tangible opportunities by brainstorming and sketching potential solutions using the Crazy Eights method.
Through collaborative sessions, we explored diverse ideas, rapidly visualized concepts, and began shaping how these solutions could address the challenges students and early-career professionals face when learning new skills.
Value Proposition
Which design concepts provide the most value to students and early-career professionals?
Conducted User Interviews to evaluate which concepts resonated most with students and early-career professionals. The goal was to test value propositions and understand how users perceive usefulness, ease of use, and professionalism across different design directions.
“The personalization by interests is great, because talking about my own skills feels intimidating.”
“AI is good for feedback, but I don’t want it doing the entire work for me.”
“I’d use something like this if it helps me stay accountable and actually apply what I’m learning.”
Final Design
AI Job Simulator
AI Mock Interview
AI Learn Flow
Simulates job tasks and projects, allowing users to practice skills in realistic scenarios
Provides real-time and post-interview feedback on tone, clarity, and answers.
Builds personalized learning paths based on user preferences.
Reflection
AI Integration
Learned how to incorporate AI into design workflows while recognizing its strengths and limitations.
User Research
Improved skills in interviewing and synthesizing insights from diverse groups across career stages.
Collaboration
Strengthened teamwork and communication when transforming findings into design solutions.
Problem-Solving
Built confidence in addressing challenges and iterating based on real user feedback.