Synapse

Inspiration

We were inspired to reimagine how organizations gather consumer insights and the intersection of biometric analysis and real-time personalization. We asked: What if feedback collection became invisible and integrated into natural digital behavior? Combined with biometric emotion detection, we saw an opportunity to capture authentic sentiment in real-time, transforming passive social media scrolling into active insight generation.

What it does

Synapse is a real-time emotion detection and behavioral analysis platform that seamlessly captures consumer sentiment through behind a social media facade. It:

• Captures authentic emotional reactions via facial recognition as users interact with content (videos, products, media) using Presage.
• Tracks emotional states (Happiness, Sadness, Anger, Surprise, Fear, Disgust) with precise strength metrics in real-time
• Harvests behavioral interests by correlating positive emotional responses with content tags and metadata
• Monitors engagement patterns through "boredom streak" detection to identify moments of peak receptiveness
• Broadcasts prime selling opportunities via the Solace message broker when emotional vulnerability peaks
• Calculates dynamic data market value based on emotional intensity and engagement state
• Provides tailored advertising strategies for each emotional state
• Suggests a custom list of potential products to purchase based on consumer interests

How we built it

• Frontend: React + TypeScript with Vite for responsive UI development
• Real-time Emotion Detection: Presage SDK for continuous facial emotion recognition without user intervention 8 Behavioral Data Harvesting: Automatic tagging and interest accumulation tied to emotional response triggers
• Real-time Messaging: Solace PubSub+ broker for broadcasting consumer insights across organizational systems
• Backend Processing: Python processor for aggregating emotion data streams and behavioral patterns
• Suggested Products: Solace agent uses Yellowcake to scrape online retailers for suggested purchases curated to the user's emotional profile.

Challenges we ran into

• Latency: Managing emotion detection latency while maintaining real-time polling frequency
• Webcam Access: Handling browser permissions and ensuring reliable video stream initialization
• False Positives: Filtering genuine engagement signals from momentary expressions

Accomplishments that we're proud of

• Eliminated Survey Friction: Created a system where insight collection happens invisibly during content consumption
• Real-time Sentiment Pipeline: Integrated Presage SDK with continuous polling to capture authentic emotional responses
• Boredom Streak Detection: Engineered a vulnerability-window system that identifies optimal moments for intervention
• Dynamic Valuation Model: Built a pricing algorithm that adjusts data market value based on emotional intensity and engagement state
• Cross-system Broadcasting: Implemented Solace message broker integration for organizational-scale data distribution
• Behavioral Profile Export: Created exportable dossiers of user interests and emotional patterns for business intelligence

What we learned

• Emotion detection APIs are powerful tools for understanding authentic consumer behavior at scale
• Interest correlation: Consumer sentiment directly correlates with content topics, enabling precise behavioral profiling

What's next for Synapse

• Multi-user Tracking: Scale biometric analysis across multiple subjects simultaneously across different platforms
• Cross-Platform Expansion: Extend beyond web to mobile apps, in-store retail environments, and smart devices
• Advertiser Network: Sell real-time emotional insights to brands for precision targeting

Built With

• c++
• docker
• presage
• python
• react
• solace
• typescript
• yellowcake