Revolutionizing Fire TV Search with Generative AI

DESIGN FOR TV

Overview

I had the pleasure of working as a UX Designer on Amazon’s FireTV team, where I focused on modernizing the Search experience. As streaming platforms continue to evolve, integrating cutting-edge technologies like Generative AI into FireTV's search became a critical step in transforming how users find content. This project aimed to create a more intuitive, responsive, and personalized search interface for over 50+ million users.

While I can't share all project details or my most in depth process, I can highlight key aspects of my role. I led UX design efforts, collaborating closely with teams across Voice design, Product management, and Research. Together, we explored innovative ways to apply AI in search, enhancing user interactions by enabling more natural, conversational experiences and quick filters.

One of my key contributions was designing quick filter pills to streamline the search process, allowing customers to refine their results faster and more efficiently. I incorporated this feature into an interactive prototype, which was tested to ensure it enhanced their overall search experience. Throughout the design process, I consistently presented proposals and refined solutions to stakeholders, ensuring alignment across teams and a clear vision for enhancing the FireTV search experience.

My Role

Designing the experience end-to-end, working closely with engineering to amplify search and discovery for customers in detail.

The Team

1 of 2 designers, with guidance from a Lead designer, 1 Voice Designer, 1 Product Manager, 8+ engineers.

Timeline

February 2021 - June 2021

💌 Get in touch for more details.

Background

The Current Search Experience on FireTV

What is Search?
Search is a core feature on FireTV, allowing customers to find the content they want to watch easily

How does Search work?
Customers can enter their queries via keyboard or through voice commands using Alexa. However, while functional, the experience has remained relatively static over time

Why the Change?
As user expectations evolve and AI continues to transform digital experiences, FireTV's traditional search system started to feel outdated. With the rise of Large Language Models (LLMs) and Generative AI, it became clear that the current search lacked the intelligence and responsiveness users were beginning to expect. This created an opportunity to bring the search experience into the future.

The Opportunity

By integrating Generative AI into FireTV's search, we aimed to revolutionize how users discover content, making searches more intuitive, conversational, and personalized. This upgrade would not only modernize the experience but also keep FireTV competitive in a rapidly advancing market.

Unlocking the Future of Search: Integrating AI on Fire TV

How might we revolutionize FireTV search by leveraging AI to deliver a more intuitive, personalized, and seamless experience that anticipates customer needs and enhances content discovery?

Why Was This Change Necessary?

Integrating AI into FireTV wasn’t just about innovation—it was a critical step in enhancing the customer experience and staying competitive in an evolving digital landscape.

Elevating the Customer Experience: With Generative AI, FireTV Search and Alexa became more conversational and intuitive, offering a natural, human-like interaction. This shift transformed the search process into a more seamless and engaging experience, allowing users to find content with greater ease.
Taming a Vast Content Library: FireTV's extensive content catalogue can be overwhelming for users. By introducing AI, we empowered customers to navigate this immense library with precision, offering personalized, context-aware suggestions that made content discovery faster and more relevant.
Leading the Future of UX Design: AI is reshaping the entire field of UX; by integrating this technology into FireTV, we positioned the platform at the forefront of innovation, pioneering the application of AI in entertainment and setting new standards for the industry.

What Were the Challenges?

Despite Amazon's significant investment in resources, the project posed unique and formidable challenges that required innovation and adaptability.

Uncharted Territory: Implementing AI in FireTV Search, especially on a TV platform, was a pioneering effort. Unlike other projects where we could draw from existing solutions, this was a completely new frontier. We weren’t just adding a feature—we were designing an entirely new search experience without a clear roadmap.
Balancing Innovation with Legacy Systems: FireTV’s well-established design system, while robust, posed challenges when introducing such a groundbreaking feature. We had to strike a delicate balance between maintaining consistency with the existing system and pushing the boundaries to create a seamless, AI-driven experience that felt fresh without disrupting familiar user interactions.
Discoverability and Customer Understanding: Introducing a feature as complex as AI required clear communication with customers. Ensuring they understood its benefits and how to use it was crucial. We focused on making the feature easily discoverable and intuitive, ensuring users could immediately grasp its value and interact with it naturally.

Customer-Centric Design: Insights That Drove Our AI Strategy

Elevating the Customer Experience with AI-Driven Search

Integrating such a powerful AI feature into an established design system required a thoughtful reimagining of the customer experience. Here’s how we transformed it:

Enabling Natural, Conversational Queries: Previously, FireTV’s search capabilities were limited to short, direct queries like "Ant Man" or "Action Movies." With AI integration, customers can now ask more detailed, natural language questions, such as "What was that superhero movie with Paul Rudd?" This shift allows for a more human and intuitive interaction with the platform.
Personalized Refinements: Beyond longer queries, customers can now refine their searches in real-time. For example, after asking about a specific movie, they can narrow results further by saying, "Now only show me the ones that are free to me." This ability to add layers of refinement creates a highly personalized and efficient search experience that wasn’t possible before.
Maintaining Design Consistency: While introducing this groundbreaking feature, we ensured the core elements of the design system remained consistent and familiar to users. Balancing innovation with the established interface, we worked within the boundaries of the existing design to make the transition as seamless as possible, keeping the learning curve minimal.

Mapping the Customer Journey: From Uncertainty to Discovery

To ensure our AI-driven search met diverse user needs, we crafted a detailed customer journey map. This allowed us to understand how customers would interact with Generative AI Search—whether they knew exactly what they wanted or had no idea where to start.

When the Customer Knows Exactly What They Want: For customers with a specific movie or show in mind, but who may have forgotten the title, Alexa can now interpret detailed descriptions. By providing clues like "that superhero movie with Paul Rudd," users can efficiently find exactly what they're looking for, even with limited information.
When the Customer Has No Idea What to Watch: For those uncertain about what to watch, Alexa guides them through a curated experience. Customers can start with a broad search, like requesting a genre, and then progressively refine it by adding preferences such as "only show family-friendly comedies." This personalized approach transforms the search into a journey of discovery, helping users uncover content they might not have found otherwise.
Dynamic Search Refinements: One of the major advancements of this project is the ability for users to build on their search queries. Each step in their search journey—whether it starts with a broad genre or a vague memory—can now be refined with additional criteria. The result is a fluid, multi-step search process that evolves with the customer’s input, leading them to their perfect match.

Design Solution: Bridging Innovation and Usability

What Did We Change: Thoughtful Solutions for Evolving Customer Needs

To address the key customer needs identified, we implemented carefully considered design changes that seamlessly integrated with FireTV’s existing system while introducing powerful new functionalities. Here’s how we aligned customer needs with innovative solutions:

Revolutionizing Longer Customer Queries and Search Sessions: The introduction of AI has transformed the way users search on FireTV. Now, users can begin with broad, open-ended queries and progressively narrow their choices through follow-up questions, creating a dynamic and conversational search experience. To support this breakthrough, we redesigned the breadcrumb in the top-left corner to seamlessly accommodate longer, more complex queries, giving users a clear visual map of their search journey. This innovation not only simplifies the process but also empowers users to refine their searches in ways that were never possible before.

Quick Filter Pills for Effortless Refinement: We introduced quick filter pills to streamline content discovery. These intuitive visual cues allow users to instantly refine their results—filtering by options like "free to me" or "new releases"—without restarting their search. This on-the-go refinement saves time and eliminates friction, making search faster and more responsive.
Balancing Innovation with Familiarity via FireTV’s Design System: While we pushed the boundaries of what search on FireTV could do, maintaining user comfort was critical. Our design solution harmoniously blended new AI-driven features with the existing interface, ensuring a familiar experience for users. Despite the powerful new features, we were mindful to preserve the familiar look and feel of FireTV Search. By utilizing existing design components, we ensured that users could enjoy the new functionality without sacrificing the consistency and ease they expect from the platform.

1.Revolutionizing Longer Customer Queries and Search Sessions

Before

Before: The breadcrumb displays the exact wording of the customer’s search query.

After

After: The breadcrumb is divided into two lines. The top line displays the exact wording of the customer's query, while the bottom line shows how the LLM interpreted it, along with any refinements from previous searches.

Why Did We Do This?

As search interactions become more conversational and complex, we anticipated that customers would begin using longer, more detailed queries. To support this shift, we designed a breadcrumb system that could handle the increased complexity while enhancing user clarity.

The First Line: Displays the exact wording of the customer's previous query, ensuring transparency and helping users confirm that the system understood them correctly. This provides reassurance that their search is progressing as intended.
The Second Line: Works in harmony with the first, dynamically tracking all refinements users make to their original query. For example, a customer might start with "Idris Elba" and later refine their search by adding, "only show the free ones" or "only show new releases." These refinements are displayed in the second line, offering a clear, organized view of how the search is evolving in real-time.

By giving users this visual feedback, we created a more intuitive, responsive search experience that adapts to their needs while making it easy to follow and refine their journey.

2.Quick Filter Pills for Effortless Refinement

Before

Before: Previously, there was no way for customers to refine their search, meaning each query stood alone and was disconnected from the last.

After

After: We introduced filter pills, empowering customers to build on their existing searches instead of starting from scratch. These filters allow users to make their searches more precise and tailored to their needs, encouraging a more fluid and personalized search experience. By enabling granular refinements, we help customers find exactly what they’re looking for faster and with greater ease.

The Reason Behind the Change

The introduction of filter pills was driven by our goal to improve discoverability and make the AI-powered search more intuitive and flexible for customers. By adding these pills, we visually signal that search refinement is possible, guiding users to explore beyond their initial query. Customers can either tap a filter or speak their refinement to Alexa, making the experience seamless across input methods.

Additionally, each refinement is tracked in the breadcrumb, offering users a clear view of how their search is evolving. This feature empowers customers to use a funnel search method, allowing them to start broad and narrow their results over time. As a result, users no longer need to know exactly what they're searching for from the start, making content discovery more natural, dynamic, and user-friendly.

3.Balancing Innovation with Familiarity via FireTV’s Design System

Before

Before: Previously, the FireTV search experience was static, relying on familiar design patterns and providing limited interaction beyond basic queries.

After

After: While the overall layout remains mostly the same, the customer experience (CX) has been transformed. This deliberate choice was made to ensure users don’t feel disoriented in the new AI-driven interface. We struck a careful balance between integrating powerful new features and maintaining the familiar design patterns users trust. This approach allows customers to explore the enhanced functionality with ease, without facing a steep learning curve or feeling disconnected from the platform they know.

The Logic Behind Our Design

Our challenge was to introduce a cutting-edge AI-driven search experience without overwhelming the customer or disrupting the familiar interface they trust. To achieve this, we carefully designed the new features using existing components from the established FireTV design system. This approach allowed us to seamlessly integrate powerful new capabilities while preserving the intuitive experience customers are accustomed to. By maintaining consistency, we ensured that users could explore the new functionality with confidence, without sacrificing ease of use or trust in the platform.

Pioneering the Future of Entertainment with AI

Designing for the new world of Generative AI, Large Language Models (LLMs), and AI-driven solutions is an exhilarating challenge and represents the next frontier in user experience design.

Working on FireTV was particularly inspiring because it consistently pushes the boundaries of both technology and entertainment. Designing AI for this space was both exciting and demanding, as we were breaking new ground with almost no precedent to follow. It forced me to think creatively and innovate new solutions to the unique challenges we faced.

Seeing the project I contributed to being showcased on the main stage of a tech conference was a dream realized, validating how far I've come in my career. It felt like a personal milestone, reflecting both my growth as a designer and the potential of what we had built together. I’m incredibly grateful to the entire FireTV team, especially my fellow UX designers, and the brilliant engineers, executives, and managers who collaborated on this journey. Together, we set a new standard for AI in entertainment; I couldn’t be more proud.