RealTalk
An AI-driven language learning app focused on real-world travel conversations, feedback, and long-term engagement.
Fall 2022
Turning listening into learning.
Overview
Role
Solo UX/UI designer
Tools
Figma/FigJam
Maze
PROBLEM
Language learning apps are everywhere. Retention, however, is not.
As more people turn to digital platforms to learn a new language—especially for travel—engagement drops drastically. Edtech apps already face some of the lowest retention rates of any mobile category, with usage declining dramatically within the first 30 days.
At the same time, language learners consistently report the same frustration:
The problem isn’t access to content. It’s the lack of authentic, low-pressure conversational practice that simulates real-world interactions.
PROBLEM
Despite spending significant time on language-learning apps, users struggle to confidently engage in conversations in their target language.
Most platforms:
Introduce vocabulary in isolation
Force users through rigid learning paths
Provide little explanation or learning value when mistakes occur
As a result, learners disengage just when they need practice the most.
MARKET LANDSCAPE
USER RESEARCH
People don’t want to just learn a language—they want to use it.
Through interviews and affinity mapping, several patterns about language learners' approach to learning a new language emerged:
Across all interviews, one truth remained consistent: learners don't just want to learn new words—they want to be able to speak the language.
OPPORTUNITY
Further research into top language apps revealed three major gaps:
Opportunities
Vocabulary taught in isolation
Teach language in real-world contexts
Rigid learning paths
Allow user-driven, goal- and topic-based learning
Mistakes penalized, not explained
Turn errors into teachable moments
This led to a central question:
How might we help learners practice meaningful conversations for travel so they stay engaged and confident enough to keep learning?
Hypothesis
Providing immersive, conversational practice with real-time feedback will improve speaking confidence and increase long-term engagement in language learning.
DESIGn STRATEGY
To support both user motivation and retention, the experience was designed using Nir Eyal’s Hooked Model, aligning behavioral psychology with learning goals.
Trigger: “I need to practice speaking before my trip.”
Action: Users engage in conversation challenges that simulate realistic travel scenarios using AI-powered dialogue.
Investment: Steady progress with effort
Variable Reward: Upon completing a challenge, users earn a Passport Stamp—a collectible symbol inspired by real travel.
Loop returns to Trigger stage
Usability Testing
92% success rate in completing key tasks.
Usability testing showed a 92% success rate in completing key tasks, with users consistently describing the experience as intuitive and approachable—especially important for a learning product.
Feedback and mentor critiques led to key refinements:
UI color updates to shift away from warning-associated tones and toward a more welcoming palette
Micro-interactions and animations to provide clearer system feedback
Scaffolded learning support, including:
Word definitions in isolation
Pronunciation guidance
Clear explanations after errors
These iterations reinforced the product’s core promise: learning through encouragement, not correction.
Solution
A conversational language-learning experience centered on travel-relevant scenarios, powered by natural language processing and designed to feel supportive, not performative.
Rather than memorizing phrases, users practice conversations they’ll actually have—ordering food, asking for directions, checking into a hotel—while receiving immediate, constructive feedback. Users can:
Practice conversations they’ll actually encounter while traveling
Receive immediate, supportive feedback using NLP
Stay motivated through progress tracking and gamified rewards
PROTOTYPE
REFLECTIONS
This was my first shot at a UX case study project! Through the support of my mentor, I learned a lot about design thinking, what questions to ask, the importance of behavioral psychology and iteration—not only on visual designs, but also on user flows and ideas—and the art of pivoting from initial ideations.
Huge shoutout to my DesignLab mentor, Cornelia!
[P.S. 2026 reflection]
Time really does fly! Looking back at this project three years later, I’m proud of the work baby-designer jenny put into it. The foundation is there—the research instincts, the curiosity, the care for users.
At the same time, distance brings clarity. With more experience in product design and a much deeper understanding of AI, I now see new possibilities I couldn’t have articulated then. Now, I’d approach the problem with sharper prioritization, tighter product framing, and a more intentional use of AI as a true learning partner rather than just a feature.
This project captures where my thinking was at the time. If I were building this today, I’d focus less on what the product could do and more on how it adapts, personalizes, and grows with its users over time.
Growth doesn’t erase earlier work—it gives it context. And this project remains a meaningful milestone in that journey.