Portfolio
Case study 02
Sharpener (also known as "Tinder for tutors") is a prototyped mobile application that matches students with educators based on specific academic needs and subject expertise.
Role
Lead UX Designer & Researcher
Timeline
6 weeks / January 2025
Tools
Figma, Miro
Scope
Academic Portfolio Project
Phase 01
Context & The Challenge
A critical drop-off problem hiding in plain sight
Background
Traditional tutoring platforms ignore learners' emotional states and specific curriculum levels. I designed Sharpener—a mobile concept using algorithmic matching—to eliminate pairing anxieties and streamline student-tutor compatibility over a 6-week research and prototyping process.
Problem Statement
Students, parents, and tutors need a mobile-first matching service focused on personality and learning styles to replace frustrating, trial-and-error search tools that hinder academic progress.
Primary KPI goal
Target reduction in user pairing abandonment, achieved by replacing manual search filtering with a progressive, bubble-based tutor onboarding flow.
Phase 02
Discovery & Empathize
Uncovering the friction in student-tutor compatibility
I analyzed existing tutoring services to map the emotional vulnerabilities of struggling learners facing imprecise matching, high costs, and academic anxiety.
Research methods
6 weeks
Project timeline duration
2 groups
Target users analyzed (Students & Tutors)
4 products
Included in the competitive audit
Core user pain points
FROM COMPETITIVE AUDIT ANALYSIS
01
Imprecise Tutor Matching
Platforms fail to account for specific curriculum levels or learning styles, leaving students paired with mismatched expertise.
02
Emotional and Financial Burdens
Trial-and-error matching creates high financial costs and acute academic anxiety for struggling learners and parents.
03
Subpar Mobile Experiences
Dense, desktop-first search filters cause cognitive overload, making on-the-go support difficult to find.
Phase 03 & 04
Define & Ideate
From Research Insights to System Architecture
How Might We
This statement emerged from a competitive audit mapping learner anxieties, aligning the project logic on a single guiding objective before wireframing.
System Architecture & Design Decisions
01
Progressive Disclosure
The flow guides users from general subjects to specific classes, preventing information overload during initial onboarding.
02
Granularity in Subject Selection
The "My subjects" screen uses a bubble-based interface to gather precise data points required for effective AI algorithm matching.
03
Structured Class Hierarchy
Targeted prompts capture exact curriculum levels (e.g., Algebra vs. Calculus), feeding the matching engine to pair student weaknesses with tutor expertise.
Phase 05 & 06
Design, Prototype & Test
Iterative evaluation: Addressing cognitive load and user anxiety
Using principles from educational psychology, I audited the AI matching flow to eliminate onboarding anxiety and minimize cognitive friction.
Iteration 01
Anxiety-mitigated onboarding entry
EVALUATION INSIGHT
Early requests for performance history made users anxious. I changed the entry flow so users could see the value of matching and pick their subjects first, before being asked for detailed background data.
Iteration 02
Bubble-based data collection
EVALUATION insight
To eliminate decision fatigue, I broke complex subject selection into visual, bite-sized steps. This progressive breakdown increased completion rates and confidence.
Iteration 03
Algorithmic pairing logic
EVALUATION insight
Surfacing active class chips and highlighted match percentages streamlined tutor comparison, making the AI pairing logic transparent and reducing decision fatigue.
1
Semester
Longitudinal study duration
HCI
Structured validation framework
2
Phases
Evaluation phases
Phase 07
Impact & Key Takeaways
Anticipated academic impact & future research directions
Because this prototype remains unverified, Phase 07 establishes an HCI framework to guide future long-term behavioral testing on anxiety reduction.
TARGET KPI
Target improvement in academic outcomes
TARGET KPI
Target confidence level for statistical significance
ANTICIPATED IMPACT
Core validation framework alignment
ANTICIPATED IMPACT
Average time-to-submit reduced to
What I’d carry into the next project
01
Design from real user behavior, not assumptions
Smart features like AI matching should always solve proven user frustrations, not just exist because the technology is available.
02
Keep the matching fair for every student
AI models must be trained on diverse student backgrounds so that everyone gets equal access to quality tutoring.
03
Routinely audit the algorithm for fairness
Responsible design requires ongoing testing to spot flaws early and make sure the system stays fair and reliable.
© 2026 Streett Jacobs — All rights reserved