PAWSE

Inspiration

Although online purchases have become easier and more accessible than ever, such level of access has led to many individuals suffering from impulsive/compulsive buying habits stemming from other mental disorders.

What it does

We built a chrome extension that automatically detects when a user accesses the checkout page of online marketplaces such as Amazon which then redirects them to a custom webpage that we also built. This website uses the webcam to take a photo of the user and scans their emotions -- redirecting them back to the checkout link if they are in a stable mood and blocking them from purchases if they are seen to be in a stressful mood.

How we built it

We used CSS and HTML to build the UI of the extension, and used JavaScript to handle the backend implementation with redirecting the user to our website. As for the website itself, we used HTML for UI and JavaScript for implementing and handling Hume API calls to accept emotion values of the user photos.

Challenges we ran into

We spent a lot of time trying to get the extension itself to access the camera and handle Hume API implementation before realizing that there was a problem on the browser end with extensions handling such functions. Due to such unforeseen limitations, we were forced to pivot our approach to redirecting the user to a custom website that would instead handle the camera input and Hume API implementation. We ran into another problem where the the real-time analysis of the camera input through websockets were not working as intended, so we opted to processing batches of media files with POST and GET requests instead.

Accomplishments that we're proud of

Some accomplishments we are proud of are building a robust Chrome extension that seamlessly intercepts Amazon’s checkout flow and redirects it to our custom emotion‐check page, integrating live webcam capture with the Hume AI Batch API to analyze user emotions in real time, and designing a clean, responsive UI. We also added a toolbar popup that tracks and displays the total number of blocked impulse purchases with a persistent storage and reset functionality. Altogether, we’ve combined advanced browser APIs, thoughtful UI/UX design, and solid state management into one cohesive tool that helps prevent impulsive purchases.

What we learned

We learned a lot of the intricacies of working with Hume API. We also learned how to create chrome extensions with JavaScript and HTML, as well as how to set up a simple local website that can handle API calls.

What's next for PAWSE

Going forward, for future development, we wish to prioritize expanding PAWSE’s reach by integrating with more e-commerce platforms beyond Amazon. We also plan to enhance its core analysis by incorporating multi-modal inputs like voice and facial recognition. Users will gain more control through configurable interventions like cool-down timers or justification prompts, and finer tuning of emotion sensitivity thresholds, moving beyond a simple purchase block.

Improving the user experience involves creating a smoother integration during checkout and providing clearer, real-time visual feedback, such as dynamic indicators to help users better perceive their emotional state and potential uncertainty. A key addition will be a user dashboard displaying intervention history, emotional triggers, and an estimated calculation of money saved by blocking impulse purchases, directly demonstrating the tool's impact.

Finally, continuous performance optimization will ensure PAWSE remains efficient. These combined efforts, broader platform support, enhanced analysis and user control, refined UX with more precise feedback and savings tracking, and robust performance, aim to make PAWSE a more versatile and insightful tool, effectively helping users make more conscious online purchasing decisions.

Built With

• api
• chrome
• css
• extension
• html
• hume
• javascript