v1.0.0-stable: Personal Document Intelligence

This initial stable release introduces the core architecture of mySearch, a high-performance RAG (Retrieval-Augmented Generation) system designed for private document intelligence. This version establishes the hybrid Java-Python pipeline and full integration with the Google Gemini 3.0 ecosystem.

Key Features:

• Real-Time Indexing: Native Java WatchService implementation for sub-millisecond document detection.
• Multi-Format Support: Robust ingestion for PDF, DOCX, TXT, MD, CSV, and Excel files.
• Semantic Retrieval: Integration with Google text-embedding-004 and FAISS for conceptual search accuracy.
• Intelligent Generation: Leverages Gemini 3.0 Flash for low-latency, context-aware responses.
• Enterprise Robustness:
  • Cross-process file locking via portalocker to prevent index corruption.
  • Recursive character text splitting for optimized context window management.
  • Exponential backoff for API rate-limit resilience.

Installation:
Please refer to the README.md for environment setup and DEVELOPMENT.md for technical architecture details.