LifeEmbedding
Cloud-native system to model human life trajectories as embeddings with retrieval-augmentation
Cloud-Native Human Trajectory Modeling
LifeEmbedding is a cloud-native system that models and visualizes human life as a trajectory embedding of events. Using retrieval-augmentation and vector databases, the system enables semantic search and analysis of life events.
Key Features
- Event Embeddings: Convert life events into dense vector representations
- Trajectory Modeling: Represent human life as continuous temporal trajectories
- Retrieval-Augmented Generation: Query similar life experiences across individuals
- Vector Database Integration: Efficient storage and retrieval using ChromaDB/Pinecone
- Cloud-Native Architecture: Deployed on Google Cloud Platform
Technical Stack
Backend: Python, FastAPI ML: Sentence Transformers, OpenAI Embeddings Database: Vector DB (ChromaDB/Pinecone) Cloud: Google Cloud Run, BigQuery Visualization: Plotly, Dash
Use Cases
- Personal life story analysis and visualization
- Finding people with similar life experiences
- Life pattern recognition and prediction
- Automated biography generation
Links
- Code: GitHub Repository
- Course Project: JHU Cloud Computing (2025)
Project Type: Course project demonstrating cloud-native ML system design Tools: GCP, Vector DBs, Sentence Transformers, FastAPI