Role
UI/UX Designer
Role
UI/UX Designer
Role
UI/UX Designer
Role
UI/UX Designer
Timeline
April 2026
Timeline
July 2025
Timeline
April 2026
Timeline
April 2026
Tools
Sketch
Tools
Sketch
Tools
Sketch
Tools
Sketch
AI-Assisted Content Curation
AI-Assisted Content Curation
AI-Assisted Content Curation
AI-Assisted Content Curation
Helping administrators create timely and localized TV content with AI-generated recommendations.
Helping administrators create timely and localized TV content with AI-generated recommendations.
Helping administrators create timely and localized TV content with AI-generated recommendations.
Helping administrators create timely and localized TV content with AI-generated recommendations.
BACKGROUND
BACKGROUND
Apps Admin is used by internal administrators to manage content categories and apps displayed on TV devices across different countries, regions, and TV model.
Apps Admin is used by internal administrators to manage content categories and apps displayed on TV devices across different countries, regions, and TV model.
Apps Admin is used by internal administrators to manage content categories and apps displayed on TV devices across different countries, regions, and TV model.
Creating a new category in a TV device typically required administrators to identify relevant events, define category titles, and manually curate app selections. As content became more localized and event-driven, this process became time-consuming and difficult to scale.
Creating a new category in a TV device typically required administrators to identify relevant events, define category titles, and manually curate app selections. As content became more localized and event-driven, this process became time-consuming and difficult to scale.
Creating a new category in a TV device typically required administrators to identify relevant events, define category titles, and manually curate app selections. As content became more localized and event-driven, this process became time-consuming and difficult to scale.
My role was to design an AI-assisted workflow that helps administrators generate category ideas and app recommendations based on target countries and date ranges, while maintaining full control over the final content.
My role was to design an AI-assisted workflow that helps administrators generate category ideas and app recommendations based on target countries and date ranges, while maintaining full control over the final content.
My role was to design an AI-assisted workflow that helps administrators generate category ideas and app recommendations based on target countries and date ranges, while maintaining full control over the final content.
CHALLANGES
CHALLANGES
Designing this feature required balancing automation with human oversight.
Designing this feature required balancing automation with human oversight.
Designing this feature required balancing automation with human oversight.
1
Reducing the effort required to identify relevant content opportunities across different countries and time periods.
1
Reducing the effort required to identify relevant content opportunities across different countries and time periods.
1
Reducing the effort required to identify relevant content opportunities across different countries and time periods.
2
Helping administrators create event-driven categories without extensive manual research.
2
Helping administrators create event-driven categories without extensive manual research.
2
Helping administrators create event-driven categories without extensive manual research.
3
Ensuring AI-generated recommendations remained easy to review and refine.
3
Ensuring AI-generated recommendations remained easy to review and refine.
3
Ensuring AI-generated recommendations remained easy to review and refine.
4
Maintaining user control over the final content displayed on TV devices.
4
Maintaining user control over the final content displayed on TV devices.
4
Maintaining user control over the final content displayed on TV devices.
LIMITATIONS
LIMITATIONS
The solution needed to fit within existing content management workflows and business requirements.
The solution needed to fit within existing content management workflows and business requirements.
The solution needed to fit within existing content management workflows and business requirements.
Content quality still required human review before publication.
Content quality still required human review before publication.
Content quality still required human review before publication.
AI-generated recommendations could not be treated as final output.
AI-generated recommendations could not be treated as final output.
AI-generated recommendations could not be treated as final output.
Application selections were limited to a maximum of 20 apps per category.
Application selections were limited to a maximum of 20 apps per category.
Application selections were limited to a maximum of 20 apps per category.
xisting manual editing workflows needed to remain available.
xisting manual editing workflows needed to remain available.
xisting manual editing workflows needed to remain available.
KEY DECISIONS
KEY DECISIONS
Token Management Table and Preview Box
Used AI to identify relevant content opportunities
Instead of requiring administrators to manually research seasonal events or trending moments, the workflow begins with two simple inputs: target countries and a date range.
AI then generates category recommendations based on events and themes that may be relevant to those regions during the selected period. This helps administrators move from content planning to execution more efficiently.
Token Management Table and Preview Box
Used AI to identify relevant content opportunities
Instead of requiring administrators to manually research seasonal events or trending moments, the workflow begins with two simple inputs: target countries and a date range.
AI then generates category recommendations based on events and themes that may be relevant to those regions during the selected period. This helps administrators move from content planning to execution more efficiently.
Token Management Table and Preview Box
Used AI to identify relevant content opportunities
Instead of requiring administrators to manually research seasonal events or trending moments, the workflow begins with two simple inputs: target countries and a date range.
AI then generates category recommendations based on events and themes that may be relevant to those regions during the selected period. This helps administrators move from content planning to execution more efficiently.
Token Management Table and Preview Box
Used AI to identify relevant content opportunities
Instead of requiring administrators to manually research seasonal events or trending moments, the workflow begins with two simple inputs: target countries and a date range.
AI then generates category recommendations based on events and themes that may be relevant to those regions during the selected period. This helps administrators move from content planning to execution more efficiently.
Save Filter Flow
Kept users in control of AI-generated recommendations
AI-generated categories are presented as suggestions rather than final decisions. Administrators can select a recommendation, edit the category title manually, or regenerate alternative suggestions when needed. This ensures AI supports decision-making without replacing human judgment.
Save Filter Flow
Kept users in control of AI-generated recommendations
AI-generated categories are presented as suggestions rather than final decisions. Administrators can select a recommendation, edit the category title manually, or regenerate alternative suggestions when needed. This ensures AI supports decision-making without replacing human judgment.
Save Filter Flow
Kept users in control of AI-generated recommendations
AI-generated categories are presented as suggestions rather than final decisions. Administrators can select a recommendation, edit the category title manually, or regenerate alternative suggestions when needed. This ensures AI supports decision-making without replacing human judgment.
Save Filter Flow
Kept users in control of AI-generated recommendations
AI-generated categories are presented as suggestions rather than final decisions. Administrators can select a recommendation, edit the category title manually, or regenerate alternative suggestions when needed. This ensures AI supports decision-making without replacing human judgment.
Save Filter Flow
Designed AI-assisted app curation instead of fully automated selection
Once a category is selected, AI generates a recommended list of applications that align with the chosen category. Administrators can review the recommendations, remove apps, add additional apps manually, or regenerate the list before publishing.
The goal was not to automate content curation entirely, but to significantly reduce the effort required to create relevant app collections.
Save Filter Flow
Designed AI-assisted app curation instead of fully automated selection
Once a category is selected, AI generates a recommended list of applications that align with the chosen category. Administrators can review the recommendations, remove apps, add additional apps manually, or regenerate the list before publishing.
The goal was not to automate content curation entirely, but to significantly reduce the effort required to create relevant app collections.
Save Filter Flow
Designed AI-assisted app curation instead of fully automated selection
Once a category is selected, AI generates a recommended list of applications that align with the chosen category. Administrators can review the recommendations, remove apps, add additional apps manually, or regenerate the list before publishing.
The goal was not to automate content curation entirely, but to significantly reduce the effort required to create relevant app collections.