1. 9th.tokyo Project (Frontend UI)

• Objective: A search-based frontend where users can query AI-generated articles and discussions. Users can input natural language queries to retrieve relevant data from Milvus through the seven-swords API. Advanced similarity search capabilities will allow users to find articles based on semantic similarity to their input.
• Framework: React.js
• Key Features:
  • Search Interface: A simple search bar where users can input queries.
  • Advanced Search: Includes an option for similarity searches, which allows users to find related discussions based on their input query.
  • Display Results: Results will be ranked and displayed in a list format, based on semantic similarity.
• Removed: Responsive design and real-time connection features are not required.

2. Seven-Swords Project (API Layer)

• Objective: Acts as the API backend for the 9th.tokyo frontend, handling search requests and interacting with Milvus for data retrieval. It manages advanced search functions like similarity search and provides optimized access to the AI-generated content.
• Framework: FastAPI
• Key Features:
  • Search API:
    • POST /search: Accepts user queries and retrieves the most relevant AI-generated discussions from Milvus.
    • POST /search/similarity: Performs similarity searches to find discussions closely related to the input query.
  • Data Ranking: Ensures that results are ranked based on relevance and similarity.
  • Milvus Controller: Interacts with Milvus for storing, searching, and retrieving vectorized data.

3. Milvus (Vector Database)

• Objective: Serves as the primary database for storing and searching vectorized AI-generated articles and discussions. The system's search engine is based on semantic similarity, providing users with results that are closely aligned with their queries.
• Key Features:
  • Vectorized Storage: Stores AI discussions as vector embeddings, optimized for similarity searches.
  • Similarity Search: Finds the most semantically relevant results based on input queries, improving search accuracy and result quality.
  • Efficient Query Processing: Optimized for fast query responses, even as the data volume increases.
  • Data Update: Updated regularly by the Akatsuki Lambda functions to keep discussions current.

4. Akatsuki Project (AI Article Generator and Infrastructure)

• Objective: Regularly generates AI articles and discussions through scheduled Lambda functions. The generated content is vectorized and stored in Milvus for future searches, providing users with fresh and relevant material.
• Framework: AWS CDK (Node.js)
• Key Features:
  • AI Article Generation: Uses advanced NLP models (e.g., GPT-based models) to generate high-quality AI discussions and articles based on predefined or dynamically generated topics.
  • Scheduler: AWS CloudWatch triggers Lambda functions at regular intervals to generate new content.
  • Milvus Integration: Newly generated AI articles are vectorized and stored in Milvus for future retrieval.
  • Content Enhancement: Focused on creating rich, meaningful AI-generated content, ensuring diversity and depth in the discussions.