
Mosaic
overview
Cross-Domain Emotional Taste Mapping System
timeline
1 week
Type
CONCEPT
skills
PRODUCT STRATEGY, UX RESEARCH, PRODUCT DESIGN
the brief
How do we help people uncover the patterns behind the things they resonate with?
1 week. 1 builder. 1 challenge: helping people make sense of seemingly unrelated preferences.
context
The things people love rarely exist in isolation…
A favourite film, a song that feels close to heart, a fashion aesthetic, or even a colour palette can often feel connected in ways that are difficult to explain. Over time, these preferences don't just reflect what people consume, they become part how they express themselves and understand who they are.
At the same time, digital platforms increasingly shape how people discover and organise their preferences. Music, films, fashion, books, and visual inspiration are separated into categories, each powered by recommendation systems designed to predict what users will engage with next.
insights
The things people are drawn to are often felt before they are understood, echoed in different media, shaped by context, and cannot be reduced to simple labels.
human taste is….
Emotional and intuitive
People recognise what resonates with them before they can explain why, making taste heavily rooted in experience. When this isn't reflected back, preferences remain felt but not understood.
Woven across domains
Taste is shaped by the quality of response rather than the object, which is why similar feelings can appear across multiple domains. When these domains are categorised, the underlying connections are lost and fragmented.
Fluid and evolving
Tastes shift over time as people move through different contexts and influences, making it an ongoing evolution. While these shifts may not always be coherent, they reflect how perception adapts to changing environments.
More than a genre
Current systems often fail to capture the full complexity of what people relate to when they reduce it to a label because people do not experience things as fixed genres, but as overlapping and sometimes conflicting expressions.
the problem
Today's platforms are designed to predict taste, not understand it.
Recommendations are generated primarily from user behaviour. Yet the way people experience taste is rarely confined to behavioural patterns alone. As a result, the emotional forces that shape their preferences remain largely invisible to the systems they use.
solution
Introducing Mosaic.
By analysing preferences across multiple forms of expressions, Mosaic translates intuitive feelings into patterns that users can explore, revisit, and build upon.
Capture signals of taste
Instead of asking users to define themselves, Mosaic collects input through visual prompts, reflection questions, and personal anchors without requiring immediate articulation.
Explore your taste over time
Mosaic builds a living map of a user's preferences over time, allowing them to revisit what they have resonated with and observe how their inclinations shift over time.
Surface hidden patterns
Mosaic reveals underlying connections that users can refine and highlights creative tensions and outliers.
beyond the core brief
These are additional deep dives, trade-offs, and user understanding that helped shape the direction of Mosaic.



WANna learn more?
There is more under the hood and I'm happy to chat about my process over a call. Reach out to me at meghnaroy.2023@gmail.com !
More works







