SkillPath

View Live Page

I worked with the Tech4Good HCI team to create Skillpath, an e‑learning platform that tackles the isolation and Week‑3 dropout pattern seen in most MOOCs. Skillpath combines AI‑driven group formation, adaptive checkpoints, and structured peer feedback to help learners “learn by doing” ,reflect together, and actually finish what they start.

Role

Role

UX/UI Designer


UX/UI Designer


Responsibilities

Team lead, user insights, AI cohort concept, full UX/hi‑fi design, user validation.

Team lead, user insights, AI cohort concept, full UX/hi‑fi design, user validation.

Time

May 2021 – Sep 2021


May 2021 – Sep 2021


Problem

Conventional e‑learning platforms were built for consumption, not connection. Learners move through static videos alone, rarely see how their progress stacks up, and struggle to find timely help when questions arise. The absence of purposeful peer groups, adaptive assessments, and on‑demand guidance leaves many feeling isolated and unsure whether they are actually mastering the material.

Solution

Skillpath replaces that one‑way, watch‑and‑hope model with an AI‑orchestrated learning ecosystem. Its Group Formation engine assembles cohorts with shared interests and comparable skill levels, creating instant study circles that motivate each other. Behind the scenes, Knowledge Assessment tools continuously analyse performance, surfacing personalised insights and actionable recommendations that let learners see—and believe in—their own progress. Meanwhile, always‑on Virtual Agents act as tireless teaching assistants: answering questions, recommending resources, and nudging learners toward their goals. Together, these three layers transform an isolating course catalogue into a dynamic, supportive path that keeps more learners engaged through completion and unlocks higher conversion for the business.

Research & Insights

Our investigation set out to understand why 55%of MOOCs learners abandoned their courses by Week 3 and what pedagogical levers could reverse that trend. We approached the question from three angles: analytics, field research, and interview with UC Santa Cruz graduate students (Master’s & PhD)

Data showed most learners quit right after Lesson 10. When we inspected the video player, we found three issues that explain the drop: no progress indicator, no easy way to ask questions, and nothing that shows classmates are present—so people feel lost and alone and leave.

We ran moderated sessions with eighteen working professionals—our core demographic—to replay their actual Week‑3 workflows. Participants repeatedly voiced three frustrations:

  • “I have no clue whether I’m ahead or behind.”

  • “I wish there were study circles; I’m the only one in my cohort still posting.”

  • “The Q&A tab feels like shouting into the void—no one answers.”

Video‑coded behaviour tallied with the interview data: moments of uncertainty produced a flurry of rewind‑and‑pause actions, then a complete exit from the lesson.

To ground our observations, we distilled findings from Smith & MacGregor (1992) and Johnson & Johnson (2009) on collaborative learning, focusing on three conditions proven to raise persistence:

  • Group formation—interest‑ and level‑aligned cohorts foster accountability.

  • Continuous knowledge assessment—frequent, low‑stakes checks increase self‑efficacy.

  • Scalable peer interaction—large groups succeed when tech mediates just‑in‑time mentoring.

How the research shaped the solution

The convergence of behavioral data, user testimony led us to three design mandates and ultimately to Skillpath’s AI‑driven pillars:

Accountability spikes when learners share goals with similar peers


Micro‑assessments boost completion more than end‑of‑module quizzes


Students in 100 + person MOOCs stay engaged when guided by chat mentors

Competitive Analysis

Market gap: The leading MOOCs don’t support real‑time collaboration

Key insight — None of the major players offers structured, in‑session collaboration.
Design implication — Skillpath can differentiate by making peer interaction, shared projects, and synchronous feedback core to the experience.

Semi‑Structured Interviews

Learners crave accountability and quick help.

“I got stuck on one concept and never went back.”
“If I had a study buddy, I’d finish the course.”
“Asking a question in a forum hours later feels pointless.”

Themes we heard

  1. Timely answers — learners want to ask questions without leaving the lesson.

  2. Visible peer progress — seeing classmates advance motivates them to keep up.

  3. Belief in collaboration — most feel teamwork improves understanding and retention.

Design implication — These themes shaped Skillpath’s three AI pillars: automatic Group Formation, continuous Knowledge Assessment, and always‑on Virtual Agents for instant support.

User Persona

We researched who uses e-learning and what they want to get the most out of an e-learning platform. We categorized our target users into three main groups:

Wisdom
Seekers

These learners use e-learning to enhance their knowledge

Job
Hunters

These are learners who take up e-learning courses to get a good job

Net
Workers

These learners want to collaborate with others


“How Might We…?”

Aligned with Sara’s goals, pain points, and the collaboration gap we uncovered

  1. HMW keep learners like Sara motivated past week 3 by making their progress and achievements unmistakably visible?
    Addresses the drop‑off cliff and Sara’s need for ongoing engagement.


  2. HMW connect Sara with a flexible, like‑minded cohort so she can learn and network even when her schedule is unpredictable?
    Targets her desire for collaboration and real‑life peer support.


  3. HMW deliver hands‑on, industry‑relevant projects and rapid feedback that translate classroom concepts into career‑ready skills?
    Speaks to her goal of gaining practical knowledge for a summer internship.


  4. HMW tailor the learning path to Sara’s skill level and interests while guiding her with just‑in‑time answers?
    Combines personalisation with the need for instant help, paving the way for AI‑driven assessment and virtual agents.

Hypothesis

AI system that plays a teacher role in e-learning platforms!

If an AI coach inside Skillpath can behave like a human instructor—diagnosing each learner’s needs, forming balanced study groups, and delivering just‑in‑time feedback—then learners will feel as supported as they would in a live classroom, remain engaged past week 3, and finish more courses.

Guided‑Learning Phases

Proposed Solution

The core pain points were the lack of personalized learning and limited opportunities for peer collaboration. To address these, I proposed a phased solution powered by AI to support both independent learning and community-driven feedback.

  • Phase 1: Students begin with individualized learning and complete their assignments independently.

  • Phase 2: After submitting their assignments by a set due date, each student is paired with a peer's submission for review. Using guided prompts, students provide structured feedback and, in turn, receive feedback from their assigned peers.

  • Phase 3: With peer insights in hand, students revise and resubmit their work—reinforcing active learning through reflection and iteration.

Individual

Learning

Group

Feedback From yourPeer

Individual + Group

Development

Wireframe

Upon explorations of various flow towards the main design, I decided to create a foundation for the main screens in each phase, to better understand the user flow and validate essential elements based on variuos feedback from eight student users.

Dashboard

Dashboard

Individual Learning
Phase 1

Individual Learning Phase 1

Manage Group
Phase 2

Manage Group Phase 2

Peer Learning
Phase 3

Peer Learning Phase 3

Style Guide

To ensure Designing Emotionally Driven e-Learning system and have consistency in the project implementation, I developed a style guide which is the single source of truth that groups all the elements that will allow the teams to design and develop a product.

High-Fidelity

Based on the proposed solution, we decided to create a foundation for the main screens in each phase, to better understand the user flow and validate essential elements.

Sara Jalilian

Product Designer

Sara Jalilian

Product Designer

Sara Jalilian

Product Designer