Synapse is an invisible, real-time consumer intelligence platform disguised as a social media interface. By pairing cutting-edge biometric emotion detection with real-time event-driven architecture, Synapse transforms passive content consumption into actionable, highly structured consumer dossiers.
Winner of the Presage Challenge at uOttaHack 8.
Traditional consumer research is fundamentally broken. Surveys suffer from self-reporting bias, while focus groups are slow, expensive, and artificial.
We asked ourselves: What if sentiment collection became entirely invisible, capturing authentic human reactions in the exact moment they occur?
By embedding biometric analysis directly into the digital experiences consumers already love, we eliminated survey friction entirely. Synapse captures raw, unfiltered emotional states in real-time, turning digital behavior into high-value data markets.
Synapse acts as a real-time behavioral harvesting engine operating silently behind a sleek social media facade.
- Continuous Biometric Polishing: Leverages the Presage SDK to track precise intensity metrics across 6 core emotional states (Happiness, Sadness, Anger, Surprise, Fear, Disgust) via the user's webcam without interrupting their browsing.
- Behavioral Interest Harvesting: Automatically correlates micro-expression spikes with content metadata tags to map exact consumer interests.
- The "Vulnerability Window" Detector: Monitors engagement drop-offs to pinpoint "boredom streaks," identifying the exact mathematical window where a consumer is most receptive to new stimuli.
- Dynamic Data Valuation: Features an algorithmic pricing engine that calculates the real-time marketplace value of a user's data payload based on emotional volatility and engagement depth.
- Event-Driven Commerce: Instantaneously broadcasts prime buying signals across the enterprise via Solace PubSub+, prompting an automated agent to scrape and serve hyper-targeted product matches.
- Frontend: High-performance, responsive UI built using React (v18+), TypeScript, and bundled with Vite for optimized assets.
- Biometric Engine: Presage SDK integrated via a continuous, low-overhead polling pipeline to analyze facial telemetry natively in the browser.
- Data Pipeline: A Python processing engine that ingests raw telemetry streams, normalizes emotional data, and computes engagement vectors.
- Event Broker: Solace PubSub+ acts as the enterprise backbone, broadcasting real-time insights across organizational systems with sub-millisecond distribution.
- Autonomous Commerce: A specialized Solace Agent utilizing Yellowcake to scrape web retailers in real-time, matching targeted inventories against the user's current emotional profile.
- The Problem: Continuous polling of high-density emotion telemetry threatened to choke the browser thread and cause noticeable interface stuttering.
- The Solution: We optimized the polling intervals and decoupled the Presage SDK rendering logic from the main UI thread, ensuring a seamless 60 FPS scrolling experience.
- The Problem: Micro-expressions like accidental grimaces, blinking, or shifting lighting conditions skewed our sentiment data.
- The Solution: We engineered a smoothing algorithm within our Python backend that filters out transient facial twitches, requiring emotional states to hit a sustained mathematical threshold to qualify as a genuine interest marker.
- The Problem: Managing asynchronous webcam permissions across varying browser sandboxes while maintaining silent, seamless integration.
- The Solution: Implemented robust asynchronous error-handling fallback states that guarantee stable initialization without interrupting the user's session.