Mila AI was originally a web-based language learning platform built around AI-powered conversations.

Rather than following a predefined curriculum, users could engage in conversations that evolved according to their interests, preferences and learning progress.

However, this adaptive experience was not clearly communicated through the product. Users were presented with multiple topics from the beginning, but the interface did not explain how those choices would shape their learning journey. The AI was central to the product's functionality, yet its ability to personalise the experience remained largely invisible.

I led the end-to-end UX and UI design of the iOS app, rethinking how users discover the product, understand their progress and continue learning over time.

Mila AI's underlying learning experience was designed to evolve with each user, but the interface did not make this process sufficiently clear.

I identified three key experience challenges.

• Personalisation wasn't tangible Users could select topics and interests, but the interface did not clearly communicate how those preferences would influence future conversations and learning opportunities.
• Learning progression lacked visibility Without a clear learning path, users had limited visibility into their progress, what they had completed and what they could expect to learn next.
• Practice wasn't clearly connected to learning needs The system could identify opportunities to revisit and practise certain concepts, but these adaptive capabilities needed to become more understandable and actionable within the experience.

Design opportunity: Make the product's intelligence visible. Instead of simply improving navigation, the opportunity was to make the underlying intelligence of the product visible through the interface.

The goal was to make Mila AI's adaptive learning model tangible throughout the user journey. Rather than presenting AI as a hidden mechanism, I designed an experience that communicates personalisation from the first interaction, makes learning progression visible and connects performance with relevant practice opportunities.

The strategy focused on three key moments: how users establish their learning preferences, how their learning path evolves over time, and how the experience helps them identify what to practise next.

The redesign shifted onboarding from broad preference selection to a more guided introduction to personalised learning.

Strategy 01 — Guided onboarding

The onboarding flow was redesigned to establish a clearer connection between user preferences and the learning experience that followed. Rather than asking for a broad set of topics in a generic way, the flow introduced personalisation as part of the onboarding narrative.

• Helped users define language goals, level and interests in a guided setup.
• Made the learning model feel relevant from the very first interaction.
• Turned personalisation from a backend mechanism into a visible product promise.

The Practice experience brings lesson review and upcoming practice opportunities into one place, helping users understand what they have already reviewed and what comes next.

Strategy 03 — Adaptive practice

Personalised learning is not only about deciding what users learn next. It also involves identifying what they may need to revisit.

Mila AI was designed to recognise when a user needed additional practice and determine how frequently certain concepts should be revisited. My design challenge was to make these recommendations actionable without adding unnecessary complexity to the experience.

• Connected practice opportunities to learning performance.
• Integrated lesson review into the broader learning experience.
• Made upcoming review activities easier to understand.
• Designed practice as part of the learning journey rather than as a disconnected feature.

Mila AI was designed around a continuous feedback loop. The user's preferences and learning activity informed an experience that could evolve over time, generating new learning units and identifying opportunities for additional practice.

The redesigned experience was evaluated through exploratory comparative usability testing with a small group of users who had previously used the web app.

During testing, users completed tasks in less time and demonstrated a better understanding of Mila AI's personalised learning proposition. These observations provided an initial indication that the redesigned experience made the product easier to navigate and its value easier to understand.

The project resulted in a mobile-first experience designed to make Mila AI's core value proposition more tangible: a language learning journey that evolves with the user.

The redesigned experience brought together guided personalisation, progressive learning paths, adaptive practice and easier session continuity. By connecting onboarding, learning progression and practice, the design aimed to make the AI's adaptive capabilities more visible and understandable throughout the user journey.