Youtube
AI Mixes
An AI-driven YouTube feature that curates mood-based playlists using user preferences and creator patterns
Youtube
AI Mixes
An AI-driven YouTube feature that curates mood-based playlists using user preferences and creator patterns
Youtube
AI Mixes
An AI-driven YouTube feature that curates mood-based playlists using user preferences and creator patterns
Objective
Personalized Content Discovery & UX Innovation
Reimagine YouTube’s content experience through AI-driven personalization that curates dynamic video mixes based on users’ moods, interests, and creator preferences. This project explores how adaptive AI can deepen engagement, reduce decision fatigue, and make content discovery more intentional and emotionally resonant.
My Role
UX Research
Concept Development
Wireframing
Prototyping
Interactive Design
Tools
Canva
Figma
Figjam
Duration
Research- 2 weeks
Prototyping- 3 weeks
Team
Esha Macha
Objective
Reimagine YouTube’s content experience through AI-driven personalization that curates dynamic video mixes based on users’ moods, interests, and creator preferences. This project explores how adaptive AI can deepen engagement, reduce decision fatigue, and make content discovery more intentional and emotionally resonant.
My Role
UX Research
Concept Development
Wireframing
Prototyping
Interactive Design
Tools
Canva
Figjam
Figma
Duration
Research-
3 weeks
Prototyping-
3 weeks
Objective
Personalized Content
Discovery & UX Innovation
Reimagine YouTube’s content experience through AI-driven personalization that curates dynamic video mixes based on users’ moods, interests, and creator preferences. This project explores how adaptive AI can deepen engagement, reduce decision fatigue, and make content discovery more intentional and emotionally resonant.
My Role
UX Research
Concept Development
Wireframing
Prototyping
Interactive Design
Tools
Canva
Figjam
Figma
Duration
Research- 8 weeks
Prototyping- Ongoing
Team
Esha Macha
The Challenge
Who are we solving for?
The YouTube Mixes experience is designed for viewers seeking more meaningful and personalized content discovery.
These users often struggle to find the right content that matches their mindset or energy level and want smarter, personalized recommendations that blend their favorite creators with the mood they’re in.
what are we solving?
Currently, YouTube’s recommendation system focuses on general viewing history rather than how users feel or what mindset they’re in, making content discovery feel impersonal and repetitive.
By introducing mood-based AI mixes, users can discover videos that match their emotions, energy, and interests, creating a more intentional and immersive viewing experience.
The Challenge
Who are we solving for?
The YouTube Mixes experience is designed for viewers seeking more meaningful and personalized content discovery.
These users often struggle to find the right content that matches their mindset or energy level and want smarter, personalized recommendations that blend their favorite creators with the mood they’re in.
what are we solving?
Currently, YouTube’s recommendation system focuses on general viewing history rather than how users feel or what mindset they’re in, making content discovery feel impersonal and repetitive.
By introducing mood-based AI mixes, users can discover videos that match their emotions, energy, and interests, creating a more intentional and immersive viewing experience.
what are the end goals?
The goal is to create a more personalized and emotionally aware YouTube experience that helps users find content aligned with their current mood or mindset.
It aims to enhance engagement, reduce decision fatigue, and make content discovery feel more meaningful and connected to each user’s daily life.
The Challenge
Who are we solving for?
The YouTube Mixes experience is designed for viewers seeking more meaningful and personalized content discovery.
These users often struggle to find the right content that matches their mindset or energy level and want smarter, personalized recommendations that blend their favorite creators with the mood they’re in.
what are we solving?
Currently, YouTube’s recommendation system focuses on general viewing history rather than how users feel or what mindset they’re in, making content discovery feel impersonal and repetitive.
By introducing mood-based AI mixes, users can discover videos that match their emotions, energy, and interests, creating a more intentional and immersive viewing experience.
Problem Statement
Problem Statement
Problem Statement
How might YouTube use AI to create smarter, mood-based content recommendations that adapt to users’ emotions, preferences, and viewing habits, helping them discover videos that truly match how they feel and what they want to experience in the moment?
How might YouTube use AI to create smarter, mood-based content recommendations that adapt to users’ emotions, preferences, and viewing habits, helping them discover videos that truly match how they feel and what they want to experience in the moment?
How might we help users communicate their mood or mindset so YouTube can generate more emotionally relevant and personalized video mixes?
Personalized Mood Recognition
How might we let users guide the AI by selecting specific creators or content styles they enjoy, ensuring the playlists feel uniquely tailored to their viewing preferences?
Creator-Driven Recommendations
How might we make it effortless for users to create, edit, and refine AI-curated mixes based on a single input like a mood, video, or favorite channel?
Seamless Playlist Creation
The Solution
The Solution
The Solution
The solution introduces Youtube Mixes, an AI-powered feature that curates personalized video playlists based on a user’s selected mood, preferred creators, and viewing habits. By allowing users to express how they feel or what mindset they’re in, the AI can generate a mix that blends their desired content into a cohesive, tailored viewing experience.
This feature gives users more control and emotional connection to their recommendations while maintaining YouTube’s familiar interface.
The solution introduces Youtube Mixes, an AI-powered feature that curates personalized video playlists based on a user’s selected mood, preferred creators, and viewing habits. By allowing users to express how they feel or what mindset they’re in, the AI can generate a mix that blends their desired content into a cohesive, tailored viewing experience.
This feature gives users more control and emotional connection to their recommendations while maintaining YouTube’s familiar interface.
The solution introduces Youtube Mixes, an AI-powered feature that curates personalized video playlists based on a user’s selected mood, preferred creators, and viewing habits. By allowing users to express how they feel or what mindset they’re in, the AI can generate a mix that blends their desired content into a cohesive, tailored viewing experience.
This feature gives users more control and emotional connection to their recommendations while maintaining YouTube’s familiar interface.
How It Works
Step 1
The user seamlessly switches to the YouTube Mixes tab. Then they input their current mood/mindset, their desired creators, and click the Generate My Mix button.
This step emphasizes seamless integration, personalization, and intuitive input design, enabling users to easily express their mood and preferences for a more tailored viewing experience
Step 1
The user seamlessly switches to the YouTube Mixes tab. Then they input their current mood/mindset, their desired creators, and click the Generate My Mix button.
This step emphasizes seamless integration, personalization, and intuitive input design, enabling users to easily express their mood and preferences for a more tailored viewing experience
Step 2
Once the users' curated mix is generated, they can chat with the embedded AI if it does not meet their expectations.
This step demonstrates adaptive, conversational feedback, empowering users to refine their mix in real time through direct interaction with the AI.
Step 2
Once the users' curated mix is generated, they can chat with the embedded AI if it does not meet their expectations.
This step demonstrates adaptive, conversational feedback, empowering users to refine their mix in real time through direct interaction with the AI.
Step 3
Once the user has chatted with the AI and perfected their mix, they can click the Done button and enjoy their curated YouTube mix generated using AI!
This step captures the completion of the journey, users finalize their AI-curated mix, and enjoy a fully personalized, mood-driven YouTube experience.
Step 3
Once the user has chatted with the AI and perfected their mix, they can click the Done button and enjoy their curated YouTube mix generated using AI!
This step captures the completion of the journey, users finalize their AI-curated mix, and enjoy a fully personalized, mood-driven YouTube experience.
The Research
The Research
The Research
To better understand user needs and identify gaps in YouTube’s current recommendation system, I conducted research focused on how people interact with video content when seeking specific moods or mindsets. By analyzing competitors, reviewing community discussions, and examining user frustrations, I uncovered key insights around personalization, control, and emotional intent. This research served as the foundation for designing a more intuitive, mood-based experience that makes content discovery feel effortless and meaningful.
To better understand user needs and identify gaps in YouTube’s current recommendation system, I conducted research focused on how people interact with video content when seeking specific moods or mindsets. By analyzing competitors, reviewing community discussions, and examining user frustrations, I uncovered key insights around personalization, control, and emotional intent. This research served as the foundation for designing a more intuitive, mood-based experience that makes content discovery feel effortless and meaningful.
To better understand user needs and identify gaps in YouTube’s current recommendation system, I conducted research focused on how people interact with video content when seeking specific moods or mindsets. By analyzing competitors, reviewing community discussions, and examining user frustrations, I uncovered key insights around personalization, control, and emotional intent. This research served as the foundation for designing a more intuitive, mood-based experience that makes content discovery feel effortless and meaningful.
Competitor Analysis
I conducted a market study to identify similar theme park and navigation apps, analyzing their features, value propositions, and limitations
While some apps offer immersive AR experiences, none combine real-time translation, multilingual accessibility, and world-specific park guidance in a single, seamless solution for international visitors.
Competitor Analysis
I conducted a market study to identify similar theme park and navigation apps, analyzing their features, value propositions, and limitations
While some apps offer immersive AR experiences, none combine real-time translation, multilingual accessibility, and world-specific park guidance in a single, seamless solution for international visitors.
Competitor Analysis
I conducted a market study to identify similar theme park and navigation apps, analyzing their features, value propositions, and limitations
While some apps offer immersive AR experiences, none combine real-time translation, multilingual accessibility, and world-specific park guidance in a single, seamless solution for international visitors.
User Quotes
To gain a deeper understanding of how users currently discover and engage with content on YouTube, I analyzed community feedback from Reddit and YouTube discussion threads. These insights highlight the most common frustrations and desires around content personalization and mood-based recommendations.
User Quotes
To gain a deeper understanding of how users currently discover and engage with content on YouTube, I analyzed community feedback from Reddit and YouTube discussion threads. These insights highlight the most common frustrations and desires around content personalization and mood-based recommendations.
User Quotes
To gain a deeper understanding of how users currently discover and engage with content on YouTube, I analyzed community feedback from Reddit and YouTube discussion threads. These insights highlight the most common frustrations and desires around content personalization and mood-based recommendations.
“My recommendations are 90% the same when I refresh — I can’t get anything new that matches how I feel.”
“My recommendations are 90% the same when I refresh — I can’t get anything new that matches how I feel.”
“My recommendations are 90% the same when I refresh — I can’t get anything new that matches how I feel.”
“I just watched one video and now my entire homepage is filled with that topic — YouTube doesn’t seem to understand mood or context.”
“I just watched one video and now my entire homepage is filled with that topic — YouTube doesn’t seem to understand mood or context.”
“I just watched one video and now my entire homepage is filled with that topic — YouTube doesn’t seem to understand mood or context.”
“Why does YouTube keep recommending content I don’t want, even after I’ve snoozed or marked it ‘not interested’?”
“Why does YouTube keep recommending content I don’t want, even after I’ve snoozed or marked it ‘not interested’?”
“Why does YouTube keep recommending content I don’t want, even after I’ve snoozed or marked it ‘not interested’?”
Insights
Users want emotional context. Existing platforms recommend based on history, not how the user feels in the moment.
Limited personalization depth. Current AI tools curate playlists but don’t let users express why they like certain content.
Discovery feels repetitive. Algorithms often recycle familiar creators, limiting exploration of new voices and styles.
Persona Development
I created two user personas to capture different viewing habits and motivations. These personas help illustrate how users with unique goals—like relaxation or focus—could benefit from a more personalized, mood-based YouTube experience.
College Student
Name
Age
Behaviors
Sofia Kim
Pain Points
Feels overwhelmed by YouTube’s endless recommendations
Finds current playlists too generic or repetitive
Spends too much time searching instead of watching
22
Watches YouTube daily while studying or relaxing.
Frequently searches for “chill,” “study,” or “focus” playlists.
Motivations
Enjoys feeling emotionally connected to her playlists.
Seeks a balance between familiar comfort and new discoveries.
Digital Marketer
Name
Age
Behaviors
Marcus Rivera
Pain Points
Algorithm often repeats the same content.
Feels recommendations ignore context (time, mood, or setting).
Wishes he could tell YouTube why he likes a video to refine future mixes.
29
Subscribes to many creators across different genres.
Likes exploring new channels but dislikes irrelevant recommendations.
Motivations
Seeks smarter discovery that aligns with his favorite creators.
Values seamless transitions between work, focus, and leisure.
Persona Development
I created two user personas to capture different viewing habits and motivations. These personas help illustrate how users with unique goals, like relaxation or focus, could benefit from a more personalized, mood-based YouTube experience.
College Student
Name
Age
Behaviors
Sofia Kim
Pain Points
Feels overwhelmed by YouTube’s endless recommendations.
Finds current playlists too generic or repetitive.
Spends too much time searching instead of watching.
22
Watches YouTube daily while studying or relaxing.
Saves favorite creators but struggles to find new videos that match her mood.
Motivations
Enjoys feeling emotionally connected to her playlists.
Seeks a balance between familiar comfort and new discoveries.
Digital Marketer
Name
Age
Behaviors
Marcus Rivera
Pain Points
Algorithm often repeats the same content.
Feels recommendations ignore context (time, mood, or setting).
Wishes he could tell YouTube why he likes a video to refine future mixes.
29
Subscribes to many creators across different genres.
Likes exploring new channels but dislikes irrelevant recommendations.
Motivations
Wants AI-curated mixes that adapt to his schedule and mood.
Seeks smarter discovery that aligns with his favorite creators.
Values seamless transitions between work, focus, and leisure.
Persona Development
I created two user personas to capture different viewing habits and motivations. These personas help illustrate how users with unique goals, like relaxation or focus, could benefit from a more personalized, mood-based YouTube experience.
College Student
Name
Age
Behaviors
Sofia Kim
Pain Points
Feels overwhelmed by YouTube’s endless recommendations.
Finds current playlists too generic or repetitive
Spends too much time searching instead of watching
22
Watches YouTube daily while studying or relaxing
Frequently searches for “chill,” “study,” or “focus” playlists
Saves favorite creators but struggles to find new videos that match her mood
Motivations
Enjoys feeling emotionally connected to her playlists.
Seeks a balance between familiar comfort and new discoveries.
Digital Marketer
Name
Age
Behaviors
Marcus Rivera
Pain Points
Algorithm often repeats the same content
Feels recommendations ignore context (time, mood, or setting)
Wishes he could tell YouTube why he likes a video to refine future mixes
29
Uses YouTube throughout his day
Subscribes to many creators across different genres.
Likes exploring new channels but dislikes irrelevant recommendations.
Motivations
Wants AI-curated mixes that adapt to his schedule and mood.
Seeks smarter discovery that aligns with his favorite creators.
User Journey Map
The user journey map outlines how a viewer interacts with YouTube Mixes from discovery to reflection, highlighting their actions, thoughts, emotions, and opportunities for a smoother, more personalized experience.
The user journey map outlines how a viewer interacts with YouTube Mixes from discovery to reflection, highlighting their actions, thoughts, emotions, and opportunities for a smoother, more personalized experience.
These low-fidelity sketches explore the initial structure and flow of the YouTube Mixes feature, focusing on layout, navigation, and how users interact with mood selection and personalized video recommendations.
Opportunity Areas
Opportunity Areas
Emotional Discovery: Enable users to express their current mood or mindset and receive personalized video mixes that match it.
Emotional Discovery: Enable users to express their current mood or mindset and receive personalized video mixes that match it.
Creator-Driven Personalization: Use AI to combine users’ favorite creators and content types into tailored playlists that feel both familiar and fresh.
Creator-Driven Personalization: Use AI to combine users’ favorite creators and content types into tailored playlists that feel both familiar and fresh.
Seamless Exploration: Design an intuitive experience where users can explore mood-based playlists without leaving the core YouTube interface.
Seamless Exploration: Design an intuitive experience where users can explore mood-based playlists without leaving the core YouTube interface.
Low-Fi Prototyping:
These low-fidelity sketches explore the initial structure and flow of the YouTube Mixes feature, focusing on layout, navigation, and how users interact with mood selection and personalized video recommendations.
Low-Fi Prototyping:
These low-fidelity sketches explore the initial structure and flow of the YouTube Mixes feature, focusing on layout, navigation, and how users interact with mood selection and personalized video recommendations.
Wireframes
These wireframes translate early sketches into a clearer visual structure, defining how users navigate through the YouTube Mixes experience, from choosing a mood to exploring AI-curated playlists.
The goal was to refine usability, maintain YouTube’s familiar layout, and visualize how personalization could feel seamlessly integrated into the existing platform.
Wireframes
These wireframes translate early sketches into a clearer visual structure, defining how users navigate through the YouTube Mixes experience, from choosing a mood to exploring AI-curated playlists.
The goal was to refine usability, maintain YouTube’s familiar layout, and visualize how personalization could feel seamlessly integrated into the existing platform.
Wireframes
These wireframes translate early sketches into a clearer visual structure, defining how users navigate through the YouTube Mixes experience, from choosing a mood to exploring AI-curated playlists.
The goal was to refine usability, maintain YouTube’s familiar layout, and visualize how personalization could feel seamlessly integrated into the existing platform.
Embeded Prototype
Explore the interactive YouTube Mixes prototype, experience how users can discover moods, create mixes, and enjoy a more personalized journey.
Explore the interactive YouTube Mixes prototype, experience how users can discover moods, create mixes, and enjoy a more personalized journey.
Explore the interactive YouTube Mixes prototype, experience how users can discover moods, create mixes, and enjoy a more personalized journey.
Final Design Solution
Takeaways:
Seamless Integration Matters
Designing within YouTube’s existing ecosystem highlighted the importance of enhancing, not reinventing, familiar user patterns to create a natural, intuitive experience.
Designing within YouTube’s existing ecosystem highlighted the importance of enhancing, not reinventing, familiar user patterns to create a natural, intuitive experience.
Personalization Drives Engagement
Users respond strongly to content that reflects their moods and preferences, reinforcing how AI can humanize digital experiences through meaningful curation.
Users respond strongly to content that reflects their moods and preferences, reinforcing how AI can humanize digital experiences through meaningful curation.
Balancing Innovation with Simplicity
