Inspiration for NovaEmotion, an Emotion-Aware Voice Commerce Assistant
Typically, online shopping assistants only concentrate on keywords or filters such as category, brand, or price. However, emotions like urgency, frustration, excitement, or financial stress have an impact on actual human purchasing decisions. For instance, a person whose phone recently broke might prefer a quick and reasonably priced replacement over the one with the most features. We were motivated to develop a system that comprehends users' desires as well as their search emotions. By combining voice interaction, emotion awareness, and intelligent recommendations, NovaEmotion seeks to make online shopping more human-centered.
What it does
NovaEmotion is an AI-driven voice shopping assistant that understands both what users want and how they feel. By speaking a request like “I urgently need a cheap phone for everyday use,” the system converts speech to text, analyzes it with Amazon Nova to identify key details such as product type, budget, usage, and emotional cues like urgency or price sensitivity. It then recommends products tailored to these needs and explains why they are a good fit, creating a personalized, intuitive shopping experience that helps reduce decision fatigue.
How we built it
NovaEmotion using a modern tech stack combined with AI services: Frontend: React for the user interface and handling voice input. Voice Processing: Browser Web Speech API to convert spoken queries into text. AI Processing: Amazon Nova to analyze queries, extract intent, detect emotional cues, and generate explanation for recommendations. Backend: Node.js with Express to manage APIs and implement recommendation logic. Database: MongoDB for storing product data and metadata. Recommendation Engine: A scoring system that ranks products based on price alignment, feature relevance, ratings, and emotional factors like urgency.
Challenges we ran into
One major challenge was consistently extracting structured information from natural language queries. This was solved by designing prompts that compelled the AI to return structured JSON outputs. Another challenge involved incorporating emotional signals into the recommendation logic without making the system overly complex. This was addressed by implementing a straightforward rule-based weighting system that adjusts product rankings based on urgency or budget sensitivity. Additionally, smoothly integrating voice input with the recommendation engine posed challenges, particularly in ensuring the system responded quickly enough to provide a smooth user experience.
Accomplishments that we're proud of
We successfully built a working prototype showcasing emotion-aware shopping recommendations. The system interprets natural voice queries, extracts meaningful context, and delivers personalized product suggestions accompanied by explanations. The emotional adaptation feature, which adjusts product rankings based on the user’s situation, is a standout achievement. This elevates the assistant from a basic voice tool to a more intelligent and empathetic commerce assistant.
What we learned
Through this project, we learned that large language models can extend beyond simple chat interfaces to serve as decision-making layers within software systems. we also recognized the importance of structured prompts, seamless API integration, and designing AI systems that effectively interact with real-world data. Additionally, we gained valuable experience in building full-stack applications that integrate voice interfaces, machine intelligence, and recommendation systems.
What's next for NovaEmotion – Emotion-Aware Voice Commerce Assistant
In the future, we plan to enhance NovaEmotion with advanced personalization and intelligence. Potential improvements include real-time integration with e-commerce APIs, multilingual voice support, and deeper emotion analysis using sentiment models. we also aim to add features like review summarization, fake review detection, and AI-driven product comparison. Ultimately, NovaEmotion could evolve into a scalable platform that delivers more human-aware and intuitive digital shopping experiences.
Log in or sign up for Devpost to join the conversation.