About Sylvia AI
A closer look at the emotion-adaptive AI companion that powers the Artificial Human experience.
Sylvia AI began as a personal challenge — to build an assistant that doesn’t just listen, but truly reacts to how someone feels. I grew up trying to understand the quieter emotions in people around me, and that curiosity slowly grew into a technical question: Can a machine sense mood and adapt its tone the way humans do?
This project is my attempt to explore that question. It became a blend of expression recognition, responsive dialogue, and personalized music generation — all guided by a simple idea: technology should meet people where they are emotionally.
1. Mood Detection
Using real-time facial analysis via face-api.js, Sylvia identifies expressions such as happiness, sadness, fear, surprise, or neutrality.
The system chooses the dominant expression with a confidence score and lets the user confirm or retry.
2. Adaptive Conversation
Once the mood is locked, Sylvia adjusts her conversational tone accordingly — warm, energetic, comforting, calm, or reflective —
creating dialogue that feels emotionally aware rather than generic.
3. Music Intelligence
Sylvia offers four mood-driven music pathways: therapeutic transitions, nostalgic soundtracks, weather-based playlists, and custom “Be Your DJ” mixes.
These are powered through curated logic and playlist-generation APIs.
4. Frontend + Deployment
The interface is built with HTML, CSS, and JavaScript, with emotion scanning rendered live through the camera stream.
The entire project runs on cloud deployment through Render.
I designed and developed the full system — from mood detection logic to Sylvia’s conversational flow and the music engine. All UI elements, animations, visual styles, transitions, and page architecture were built manually without templates.
Every part of Sylvia — her voice, her reactivity, the experience around her — is the result of many iterations, experiments, late-night debugging sessions, and a lot of coffee.
I plan to expand Sylvia with voice-based emotion sensing, better memory, deeper personalization, and a more nuanced emotional model that can react to subtle changes in tone and facial micro-expressions.
Sylvia is still early — but she’s a step toward the kind of emotionally-aware AI I hope to build in the future.