Non-native speakers often feel comfortable with formal English but stumble when slang enters the conversation. A casual “That’s lit!” or a meme reference can quickly turn into confusion or embarrassment. While existing language-learning apps are strong at teaching grammar and vocabulary, they rarely address the cultural expressions that make learners feel included in everyday social interactions. This gap leaves many learners hesitant to join in, reinforcing the feeling that they are outsiders in conversations where belonging matters most.
Before jumping into design, I reviewed analogous products and learning science principles to ground our direction. Tools like Duolingo Max, Memrise, and Urban Dictionary each had strengths, but none gave learners a way to actually practice slang in context. That gap quickly became the opportunity for LingoBuddy.
At the same time, I leaned on learning science to guide our approach. Instead of overwhelming learners with flashcards or long definitions, I wanted a design that felt contextual, supportive, and personal to each learner.
Together, these insights shaped our design challenge:
What if learning slang could feel less like a test and more like a conversation?
At the start, our team had ambitious ideas for how slang learning could work — but given the short project timeline, we needed to scope carefully. We prioritized features that would deliver the most value: onboarding to personalize practice, chat to ground it in real conversations, fine-tuning to give learners control, and lightweight explanations for slang. We sketched these flows to align on scope before moving into mid-fi prototypes.
The mid-fi prototype became the basis for usability testing. Learners could click through onboarding, chat, and tone settings — and that’s when cracks appeared. Slang terms blended into the chat, making it unclear what to focus on, and fine-tuning felt repetitive and confusing when adjusted per message.
I then refined the design based on feedback: onboarding offered clearer cultural group categories, slang was visually highlighted in chat, fine-tuning shifted to a persistent top-right setting, and emoji/meme explanations were simplified. These changes made the flow smoother, more intuitive, and more aligned with how learners expected to interact.
LingoBuddy’s final design ties every feature to learning science so learning feels relevant, motivating, and memorable.
Learners set cultural background during onboarding and choose who they’re chatting with (friend, boss, colleague). This keeps learning identity-aligned and situation-specific.
Four simple tone sliders let learners shape replies without overwhelm; sensible defaults mean they can skip it.
Instead of drills, learners bring real situations (“My friend said X—how do I respond?”). The system weaves slang into suggested replies and emphasizes the slang so learners can’t miss it, turning authentic scenarios into teachable moments.
To deepen learning, slang terms can be expanded into a multimedia explanation. Learners can listen to pronunciation, view a meme for cultural context, and watch a short clip to see how the slang is used in everyday conversation.
The biggest impact of LingoBuddy was demonstrating that AI could support slang learning in a way that feels authentic, contextual, and playful. Our design embodied principles from learning science — personalization, autonomy, contextual learning, and multimedia support — and testing validated that this approach was not only feasible, but also motivating for learners.
What began as an exploration of a gap in language-learning tools became proof that slang can be practiced in conversations, not just flashcards. Even in a short course timeline, LingoBuddy showed that research-informed design and AI technology together can open new possibilities for how people learn the cultural language of belonging.
LingoBuddy wasn’t a polished product, but even as a prototype it gave me the chance to practice what I love: making interactions feel smooth, approachable, and confidence-building. Prototyping the flows and iterating through feedback pushed me to focus on details — like clarity in chat or the way settings are surfaced — that can make or break the experience for learners.
What stayed with me most, though, was the cultural side. A single label in onboarding — “cultural background” — sparked strong reactions. It reminded me how much weight small words carry, and how easily they can shape whether someone feels included or othered.
That experience confirmed the kind of work I want to keep pursuing: designing tools that help people learn, while also making them feel a sense of belonging. LingoBuddy may have been a short course project, but it pointed me toward the area I want to keep exploring as a designer — learning, culture, and connection.
