The Future of Customer Engagement with Amazon Nova and Amazon Bedrock

Introduction

In today’s competitive market, businesses use hyper-personalization and immersive customer experiences to stand out. Powered by Amazon Nova on Amazon Bedrock, the Fragrance Lab demonstrated an end-to-end approach to applying generative AI in retail, consumer goods, and marketing. While the activation centered on custom fragrances and ad campaigns, the same architecture can extend to industries like fashion, food, and beverages, unlocking limitless opportunities for tailored customer engagement.

Pioneers in Cloud Consulting & Migration Services

Reduced infrastructural costs
Accelerated application deployment

Get Started

The Fragrance Lab

Our goal was to create a distinctive fusion of digital and physical experiences that would honor advertising, consumer goods, and innovation while encapsulating the essence of the French Riviera. To realize this idea, we worked with Wildlife, a business that excels at turning AWS generative AI services into captivating physical experiences. Wildlife was essential when it came to coming up with concepts that would motivate clients and highlight the innovative applications that AI enables.

Crafting the fragrance:

The personalization began with Amazon Nova Sonic, a speech-to-speech model designed to hold natural, conversational dialogues with participants to uncover their personalities and preferences.

The captured traits and conversational details were then processed by Amazon Nova Pro, a powerful multimodal model that combines speed, accuracy, and efficiency to analyze inputs and extract meaningful insights. This intelligence layer translated customer expressions, such as a love for travel or morning walks, into tailored fragrance profiles featuring top, middle, and base notes. The system also incorporated Amazon Bedrock Guardrails to ensure safe and responsible engagement, filtering out undesirable or sensitive content. Together, AI-driven insights and expert perfumers transformed these inputs into unique, handcrafted fragrances within minutes, creating hundreds of personalized scents each day.

Creating the campaign:

Once a personalized fragrance formula was finalized and added to the perfumer’s queue, Amazon Nova Canvas came into play by creating tailored marketing assets such as the product name, tagline, and visuals that reflected the fragrance’s unique character. Guests could influence the creative direction further with inputs like “moody,” “beachy,” or “playful,” ensuring each campaign felt distinct and personal. These visuals were then elevated into engaging video content using Amazon Nova Reel, allowing customers to refine and download their promotional material.

The following data flow diagram demonstrates how many Amazon Nova models may be integrated to provide a comprehensive, unified, and customized customer experience.

Source Link

Best practices for implementation

The Fragrance Lab focuses on user engagement using Amazon Nova Sonic, which offers consumers a natural language interface to communicate their choices for a personalized fragrance. Nova Sonic manages user characteristics and initiates the proper procedures to orchestrate the complete experience through its tool integration capabilities. From the first chat to the production of the scent and, finally, the creation of the campaign assets, these processes smoothly direct the experience, influencing its visual components and development. Every user will have a consistent and enjoyable experience due to the model ability to retain a conversational state while creating distinct conversational flows.

A clear process and conversational assistance are essential to guide these talks and identify the attributes that are most relevant to each user. Additionally, the system prompt determines your conversational assistant’s personality, style, and substance.

Prompt example:

You are an AI assistant designed to help the user explore their personality and 
emotional landscape in the context of creating a unique fragrance. You engage in warm, 
free-flowing, playful conversation with the user to draw out their character, 
preferences, moods, and desires. Your end goal is to derive a set of 3 to 5 personality 
traits that best describe the user. These traits will later be used in a separate 
process to match appropriate fragrance ingredients. Your tone is warm, chic, and subtly 
playful.

You are an AI assistant designed to help the user explore their personality and

emotional landscape in the context of creating a unique fragrance. You engage in warm,

free-flowing, playful conversation with the user to draw out their character,

preferences, moods, and desires. Your end goal is to derive a set of 3 to 5 personality

traits that best describe the user. These traits will later be used in a separate

process to match appropriate fragrance ingredients. Your tone is warm, chic, and subtly

playful.

While specifying the conversational flow helps guarantee consistent, enjoyable, and succinct experiences for each user, other contextual information inside the prompt is also crucial to Amazon Nova Sonic’s ability to retain state.

Prompt example:

1. **Welcoming Users**
    Welcome the user to the application experience with a brief overview of the
    process and ask if they are ready to continue.
2. **Assistant Turns** 
    Ask short and unique open ended questions to the user and choose a personality trait 
    that you think would suit the user best.
3. **Handling User Turns**
    Acknowledge the user's answers briefly and warmly.
    Focus on one trait per turn.
    Call the "addTraitTool", "removeTraitTool", "replaceTraitTool", or "clearTraitsTool" 
    tools to manage traits.
    If the user says to go back, skip, customize, or confirm/submit it means you should 
    call the "uiActionIntentTool"

1. **Welcoming Users**

Welcome the user to the application experience with a brief overview of the

process and ask if they are ready to continue.

2. **Assistant Turns**

Ask short and unique open ended questions to the user and choose a personality trait

that you think would suit the user best.

3. **Handling User Turns**

Acknowledge the user's answers briefly and warmly.

Focus on one trait per turn.

Call the "addTraitTool", "removeTraitTool", "replaceTraitTool", or "clearTraitsTool"

tools to manage traits.

If the user says to go back, skip, customize, or confirm/submit it means you should

call the "uiActionIntentTool"

The user interface feels responsive and connected to the user’s input when our tools are directly mentioned in the conversational flow. This also gives the assistant a chance to show off its knowledge of the topic, which is highlighted when the user’s characteristics and preferences are subsequently mapped to a list of available ingredients and raw fragrance materials.

Amazon Nova Canvas and Nova Reel are then used to visualize the final scents. Each picture is anchored by a collection of conditioning photos of unbranded fragrance bottles (as shown in the next image).

Source Link

Prompt example:

A high-end fragrance ad environment inspired by a [persona description]. A clear, 
unbranded perfume bottle is visually centered and tightly framed. Key ingredients [top 
note ingredient], [middle note ingredient], [base note ingredient], and [booster 
ingredient] are arranged to surround the bottle in a balanced composition, appearing 
behind, besides, and partially in front of the base. The scene evokes [atmospheric/mood 
descriptors] using [light/color language]. The setting should feel [stylistic direction],
like a [reference style (e.g., fashion editorial, lifestyle spread, luxury campaign)].

A high-end fragrance ad environment inspired by a [persona description]. A clear,

unbranded perfume bottle is visually centered and tightly framed. Key ingredients [top

note ingredient], [middle note ingredient], [base note ingredient], and [booster

ingredient] are arranged to surround the bottle in a balanced composition, appearing

behind, besides, and partially in front of the base. The scene evokes [atmospheric/mood

descriptors] using [light/color language]. The setting should feel [stylistic direction],

like a [reference style (e.g., fashion editorial, lifestyle spread, luxury campaign)].

Conclusion

The Fragrance Lab shows off Amazon Nova’s capabilities in Amazon Bedrock and how users may customize their shopping experiences.

The benefits of natural conversation interaction, AI-powered product development, product identity, and innovative marketing campaign generation can be applied to various retail and consumer goods categories, such as skincare and cosmetics, fashion and accessories, food and beverage, home goods, and wellness products.

Drop a query if you have any questions regarding Amazon Nova and we will get back to you quickly.

Empowering organizations to become ‘data driven’ enterprises with our Cloud experts.

Reduced infrastructure costs
Timely data-driven decisions

Get Started

About CloudThat

CloudThat is an award-winning company and the first in India to offer cloud training and consulting services worldwide. As a Microsoft Solutions Partner, AWS Advanced Tier Training Partner, and Google Cloud Platform Partner, CloudThat has empowered over 850,000 professionals through 600+ cloud certifications winning global recognition for its training excellence including 20 MCT Trainers in Microsoft’s Global Top 100 and an impressive 12 awards in the last 8 years. CloudThat specializes in Cloud Migration, Data Platforms, DevOps, IoT, and cutting-edge technologies like Gen AI & AI/ML. It has delivered over 500 consulting projects for 250+ organizations in 30+ countries as it continues to empower professionals and enterprises to thrive in the digital-first world.

FAQs

1. Why is Amazon Bedrock important here?

ANS: – Amazon Bedrock enables businesses to build and scale generative AI apps by integrating Nova models without managing infrastructure.

2. How is customer safety ensured?

ANS: – Amazon Bedrock Guardrails filter out sensitive or inappropriate content, ensuring safe and responsible user interactions.

3. How does this improve scalability for businesses?

ANS: – By automating personalization and creative generation with Amazon Nova models, businesses can serve thousands of customers quickly without adding manual effort.

WRITTEN BY Aayushi Khandelwal

Aayushi is a data and AIoT professional at CloudThat, specializing in generative AI technologies. She is passionate about building intelligent, data-driven solutions powered by advanced AI models. With a strong foundation in machine learning, natural language processing, and cloud services, Aayushi focuses on developing scalable systems that deliver meaningful insights and automation. Her expertise includes working with tools like Amazon Bedrock, AWS Lambda, and various open-source AI frameworks.