From Lessons to Conversations - A Language Learning Platform Built for Real Life
Belingual is a real life client project I worked on as a Product Designer. I designed a language learning platform to help students improve their English speaking skills while building a sense of community with others on the same journey. I explored how digital tools can better support both language development and peer connection, and to create an experience that encourages real-world application.
Brief
BeLingual needed a more cohesive platform centred on the student experience—one that not only supports language learning but also fosters a sense of community among students learning together. A key challenge was enabling meaningful interaction, collaboration, and peer support, while still providing access to essential learning tools and resources.
Problem
For many students, practicing a new language felt not just challenging, but artificial. Early insights revealed that while AI-driven tools exposed learners to new vocabulary, these interactions lacked the nuance of real conversation, making it difficult to translate learning into natural speaking situations. As a result, progress in recognition and recall rarely translated into the confidence to hold an actual conversation.
Goals
Enable real, peer - to - peer conversations through scheduled speaking sessions.
Build a strong student community through shared learning experiences.
Support progress with structured tools while prioritising real world speaking engagement.
Market, Stakeholder & User Research
In this first step I wanted to gather as much information as possible concerning the market, the business and potential users to ensure every decision made and idea created was well informed. I used a series of research methods including desk research to collect secondary data concerning the market. I collated the data into a competitor analysis table which I was then able to draw insights from.
Table Insights
Insight 1: Most platforms prioritise structured or AI-led learning over real conversation. Even when speaking tasks are included it is often tutor - led and guided rather than peer - based. Why This Matters: This highlights a clear gap in the market for authentic, peer-to-peer conversation. Designing for real human interaction directly addresses a key unmet need and creates a strong differentiator.
Insight 2: Community features exist, but are not central to the learning experience. Some platforms include social elements, but these are often secondary features rather than core to the product experience. They’re either unstructured or lightly integrated. Why This Matters: There’s an opportunity to position community as a primary value driver, not an add-on. A more intentional, structured community experience can improve engagement, accountability, and learning outcomes.
Competitor Analysis Table
Stakeholder Interview
During stakeholder interviews, we explored the current state of the business and its future ambitions, aiming to understand both immediate needs and long-term direction. These conversations helped define how the platform should evolve alongside the company, while also surfacing key objectives and measurable indicators of success.
Stakeholder Objectives:
To grow into a B2B whiles keeping students at the centre
Focus on Latin American students and making the platform more personal, interactive and different from competitors.
To foster community and real relationships between students
Success Metrics:
Increased student engagement, particularly in peer-to-peer conversations and community interactions.
Improved retention rates, with students consistently returning to practice even outside of structured learning times.
User Interview
I also conducted user interviews to better understand users’ needs, preferences, and pain points with both BeLingual’s current setup and other language learning platforms. Insights revealed that the quality of teaching itself was not the issue; rather, the key challenge was applying what they had learned to real-life conversations, particularly in practical settings such as the workplace.
The User And Their Problem
After gathering all the information I started to make sense of everything by synthesizing the data. The method I used to do this was affinity mapping. The affinity map helped me to address the most common behaviours, problems and desires of the users, which I then prioritised moving forward. This then helped me create a user story, persona, experience map, problem statement and hypothesis.
Affinity Map Insights
Insight 1: Users preferred language-learning apps that used a gamified model, as the positive reinforcement helped them stay consistent in their learning and made the experience more enjoyable.
Insight 2: However, users noticed that what they learned in these apps was not always applicable in real conversations. Because of this lack of real-life context, many felt the apps were a waste of time.
Insight 3: Additionally, many users found that even if they understood a language while learning it in an app, they struggled to understand people with different accents, despite those people speaking a language they technically knew.
Persona
Experience Map
Experience Map Insights
Insight 1: The Experience Map helped me to identify where in the user journey the user was the least satisfied, which I identified as when using AI for speaking practice on another platform. This insight later informed my decision to not include AI in the platform but instead include a feature that encourages more real life conversations between peers.
User Story
Story 1: A call centre employee, I would like to communicate naturally with my clients so that I can express myself with more confidence and coherence.
Story 2: As a call centre employee, I want to practice speaking English in a relaxed way, so that I feel more confident when talking to customers.
Problem Statement
Elisa, a striving communicator, needs English learning that’s aligned with her workplace communication because general courses don’t help her feel prepared for job-specific situations.
Hypothesis
We believe creating a platform that uses immersive features and the opportunity to learn with native English speakers will equip users like Elisa to communicate better in the workplace and feel confident whiles at work. We will know this when we see better scores in her learning performance and she leaves a good review.
The Solution
During the development stage I used ideation techniques like Crazy 8’s and How Might We statements with the rest of the design team to generate as many ideas as possible which we then filtered to the most research backed and feasible ideas - Peer Chat was one of the ideas we decided to move forward with.
Peer Chat Feature
The purpose of this feature is to facilitate real-life conversations between students outside of scheduled lessons by allowing them to browse other students within the network and schedule a time and date for a call. This feature encouraged students to build friendships while practicing a language they were learning and allowed them to choose conversation topics, though they were free to discuss other subjects as well.
Design Decisions & Trade Offs
From a product perspective, we considered designing an in-built video calling feature to support real-time conversations. However, this introduced significant technical complexity, including the need for reliable video and voice infrastructure, as well as ongoing maintenance and moderation. Given that established platforms like Zoom already solve this effectively, we instead focused on creating a space for students to discover one another and have conversations externally. This allowed us to reduce development overhead while still enabling meaningful peer-to-peer interaction.
Now for the practical part…
Moving forward with the Peer Chat feature, we started of focusing on the information architecture of the entire platform. We used methods like card sorting to determine the structure and overall organisation of the platform before moving onto user flows which outlined every step the user will take to perform a desirable action. The user flow informed the wireframes created.
User Flow
Paper Wireframe
*Peer Chat Feature: Page 1*
*Peer Chat Feature: Page 2*
*Peer Chat Feature: Page 3*
Testing & Iterations
Usability Testing
I carried out usability testing on the low-fidelity prototypes with the same participants from the user interview stage.
This allowed me to observe how users interacted with the platform in a realistic context and identify any issues before moving toward the final product. To achieve this, I provided each participant with a clear, realistic task scenario designed to produce a specific outcome.
The insights gained from these tests directly informed the iterations made in the high-fidelity prototype.
Task Scenario
“You are interested in improving your English language skills and would like to set up a conversation with another student. You have time on your lunch break on Monday. Show me how you would do this.”
Users Responses
The responses we got from user testing are as follows:
“If you want to promote ‘English Around the World’ and ‘Peer Chat’, I think they should look like something you can click rather than just a piece of text.”
“Do I really need to click ‘Continue’ every time? Can’t it just move to the next step once I make a selection? And if I choose the wrong one, I’d like an option to go back.”
Users would get frustrated when they got to the summary page and have to make changes by clicking through the whole process again.
“Shouldn’t the AI assistant be sticky or pinned to the top? That’s how most platforms do it now.”
*Peer Chat Feature: Select Topic*
*Peer Chat Feature: View Partners*
*Peer Chat Feature: Summary*
Final Prototype
Reflection
Possible Impact
For students, this feature bridges the gap between learning and real-world use by enabling natural conversations. It helps build confidence, improve fluency, and fosters a sense of community, increasing motivation and consistency.
This feature can increase engagement and retention by encouraging regular use through peer connections and scheduled conversations. It also differentiates the product as a community-driven experience, supporting organic growth and reducing reliance on paid acquisition.
This shifts the focus from AI-led learning to human interaction, positioning the product as a more social, experience-driven alternative. It helps bridge the gap between structured learning and real-world application, responding to demand for more authentic learning experiences.
What I would like to add…
I believe the platform we developed has strong potential—that enhanced the learning experience for students. If given more time, I would have explored the tutor’s side of the platform which would have included a feature for attendance, grading, setting homework, individual student performance review and overall class performance review.
I also would have introduced machine learning AI to support teachers with marking and feedback. For example, an AI assistant could analyze student performance data to recommend focus areas for upcoming lessons, and even generate tailored lesson plans and progress timelines for tutors to use.
Key Takeaways…
This was my first time working on a real client project, and I’ve taken so much away from the experience. One of the most valuable lessons I learned is that users must come first. While technical feasibility and business viability are essential considerations, human desirability outweighs them. A design is only truly worthwhile if users are at the center of it.
Overall, I’m proud of the final product. It not only meets the business requirements but also addresses user needs and concerns—making it both functional and meaningful.
View Next…
Luma
Role: Product designer
Product: Luma is a beauty service booking platform designed to help clients confidently discover, evaluate, and book trusted service providers.
Duration: 2 weeks
Tools: Figma
