moodlog.io is a journaling app developed by an individual with BPD and background in CS, to help track emotions and journal in a low friction-way.
Built on modern transformer models and established psychological frameworks.
Affective computing core for emotional signal modeling
Transformer-based inference using fine-tuned RoBERTa for emotion detection
Emotion modeling structured around Plutchik's Wheel of Emotions
Pattern detection using modern NLP and NERD for contextual insights
Privacy-preserving ML architecture with local-first processing
Prosocial support features for trusted peer connections
moodlog.io does not diagnose or replace therapy. It supports reflection and emotional understanding using research-informed methods.
Privacy-first. User-owned data. No gamification of distress.