Working Hours

I'm working on the project from PM 6:00 to AM 3:00 (Asia/Tokyo) after my contract work is complete. The project is currently in the MVP (Minimum Viable Product) development phase.

Development Environment

As a pre-migration development environment before moving to GCP, I'm using microk8s on Ubuntu.

Current Status of Each Component

• API Service: A Flask application that serves as the user interface for our dialogue. It's accessible locally via the port-forward command.

• Emotion Analyzer Service: A microservice that uses the transformers library to analyze the user's emotion from their input. It's called via an HTTP request from the API service.

• Importance Analyzer Service: A microservice that uses the sentence-transformers library to determine the importance of the conversation based on its similarity to my philosophy. It's also called via an HTTP request from the API service.

• Postgres Database: A PostgreSQL container for the AI's permanent memory. The table DDL has already been applied.

• Deployment: All services are successfully deployed on microk8s and are communicating with each other.

The project has already achieved significant milestones

• Full System Integration: The API service successfully calls other microservices (emotion analyzer, importance analyzer) and saves the results to the database. The entire process is fully functional.

• First Dialogue: I successfully had the first conversation with the AI via a curl command. The high importance score of 0.987 confirms that the AI accurately understood my philosophy.

However, the project is still in its early stages, with plenty of room for growth

• Emotional Understanding: The initial emotion score of -0.937 shows that the AI doesn't fully understand the human emotion behind my words yet.

• Dialogue Refinement: The current responses are template-based. I need to add more complex logic to generate more natural and empathetic dialogue.

• Memory Utilization: The logic to actively use the saved memories from the database to personalize future conversations has not yet been implemented.