GenieMates

Project image 1
Purpose
Academic
w/Sketchin
Course
Designing Digital Experiences
Type
Group project
Role
UX, UI Designer
Year
2024
Duration
3 months

Abstract

GenieMates is a mobile platform designed to connect motivated freelancers who are eager to collaboratively learn and develop new professional skills through small, interactive group sessions.

The application integrates advanced Generative AI capabilities to enhance the learning experience. When participants encounter conversational challenges or feel momentum stalling, the embedded Generative AI intervenes as a real group member by providing constructive suggestions, feedbacks, discussion prompts to reignite dialogue and learning engagement.

The Challenge

Learning-by-doing for impact-driven solopreneurs. Develop a new digital solution (such as an app) that supports impact-driven professionals in acquiring the skills they need to build and sustain a solo business within a busy schedule.

Challenge Visual

The Solution

We designed a digital platform built on the belief (confirmed by the research) that its true value lies not in the technology itself, but in the human relationships it helps inspire, nurture and sustain. Rather than offering a purely technological solution, our focus is on enabling meaningful person-to-person connections.

As found by the research, Generative AI should serve as a quiet enabler, enhancing feedback, facilitating dialogue and supporting interaction, never replacing the human connection at the heart of it all.

AI-driven group suggestions

Intelligent group suggestions based on skills, interests, and learning goals.

Icebreaker

The AI acts as a facilitator, helping to break the ice and engage participants in the conversation, proposing topics to discuss or activities to do, such as quizzes or icebreakers.

Feedback System

The AI reinforces peer feedback within the group by reminding users to give feedback or suggesting when to ask for it.

The Process

1

User Research

Desk Research

Peer learning platforms analysis:

  • Huddelcraft Small groups, mix of in-person and online, buddy, final showcase event
  • Focusmate Having someone to do your tasks with and report to
  • Midjourney Everyone work in one workspace, get inspiration, learn from what the other are doing
  • Lunchclub Schedule weekly meeting online or in person with "unknown" peers

Main challenges in terms of scalability, quality control, and user engagement.


Key Insights

Many freelancers and solopreneurs encounter significant challenges when it comes to locating peer groups that can facilitate their learning and skill development. The nature of freelance work often leads to isolation, making it difficult to connect with others who share similar professional interests and aspirations. Traditional learning methods often lack flexibility and can involve lengthy, intense schedules. In contrast, peer learning significantly helps in achieving this learning objectives.

Research Insights

Experience Map


Learning a new skill involves a series of iterative steps. It begins with recognizing a limitation in a particular area of expertise. This leads to the discovery phase, where key questions emerge: "How can I learn this skill?", "Where can I find reputable courses?", and "How can I learn it as quickly as possible?"

  • Learning Phase Once the right course is identified, the learning phase commences. This phase consists of two distinct yet iterative components: passive learning and active practice.
  • Feedback Phase At each iteration, a feedback phase is essential. This feedback can come from various sources, such as tests, quizzes, or ideally, from real users who provide insights into your progress.
  • Integration & Maintenance After achieving a solid level of confidence in the skill, it can be integrated into your work routine. However, it's important to continuously maintain and enhance this skill over time.

Primary Research

This process led us to focus on three key findings:

  • Learning by doing Hands-on approach to skill development and knowledge acquisition
  • Peer learning Collaborative learning through shared experiences and knowledge exchange
  • Generative AI Leveraging AI to enhance and support the learning process

To further explore these areas, we conducted interviews with freelancers to understand how they balance learning activities with their daily work. Additionally, we found it enlightening to interview a teacher, providing a perspective from the other side of our research. It was particularly interesting to learn how she utilizes generative AI to assist students in learning a new language.



Key Insights

Primary Research

The research has been focused on two main aspects:

  • Peer Learning Group dimension: Group of max 8 people allow conversation where everyone is able to speak and be heard

    Vulnerability: Finding a secure environment that allows to share difficulties without the fear of being judged

    Trust: Honesty, consistency, empathy and respect along shared values and goals are the keys element of trust environment
  • Generative AI Limited way of use

    Not awareness of the potential

    limited curiosity in exploring its capabilities
2

End-to-end experience definition

Profile Building Flow

Profile Building Flow

Group Suggestion Flow

Group Suggestion Flow

Icebreaker Flow

Icebreaker Flow

Group Chat Flow

Group Chat Flow

3

UI Ideation and Design

The proposed solution is designed to be supportive, user-friendly, lightweight, and inspiring. It combines technology with human interaction. The name is a blend of "Genie" (GenAI) Mates" (Peers), representing the fusion of technology and human connection.

Design System Guidelines

Design System



Screens examples

UI Design 1

Group Suggestions

UI Design 2

Icebreaker example

UI Design 3

Feedback example



4

User Testing and Iteration

The tests we carried out concentrated on three key aspects of the app:

  • Test #1: Profile Building Feature Implementation of an autofill system to enhance the profile creation process
  • Test #2: Group Suggestion Feature Suggestions for groups accompanied by filtering options
  • Test #3: Joining a Group Evaluation of the usability involved in joining small, focused groups

Unfortunately, we were unable to test the chat feedback features and icebreaker activities due to limitations in the prototype. We utilized simulated chat interactions and required a more advanced coded prototype for comprehensive testing.

Test Findings

Test #1: Profile Building
  • AI Role Unclear AI role at the start and whether some aspects of the autofill feature are powered by AI
  • Profile Customization Give possibility to modify/add details of the profile. Some graphical elements do not convey the impression that they are editable by the user
  • Social Connections Match only relevant social connections. It's an app designed for professionals. Select the most appropriate social network to collect information
Test #2: Group Suggestion
  • Group Size Unclear limits: "What if more than 8 people want to join a group?"
  • Group Structure Deeper dive into group structure: "How user interact within the chat?"
Test #3: Chat Interaction
  • AI vs Human Interaction Initial chats felt like talking to AI, not real people
  • Progress Tracking Unclear how the progress bar reflects the quality of connections


Regrettably, because of time limitations and a restricted number of testers, we were unable to fully validate the findings or implement all the requested changes.
However, we did address some graphical elements, such as labels, component positioning, and minor user interaction features.

Test Image 1

Before

Test Image 2

After

Test Image 1

Before

Test Image 2

After

Zoomed image