EOL Framework

🚀 Performance Metrics

📊 View Detailed Performance Benchmarks

Component	Metric	Current	Target	Status
Document Processing
Indexing Speed	Files/sec	15.3	>10	✅
Chunk Processing	Chunks/sec	48.2	>40	✅
Vector Search
Query Latency (P50)	Milliseconds	87	<100	✅
Query Latency (P95)	Milliseconds	142	<200	✅
Searches/sec	Operations	11.5	>10	✅
Semantic Cache
Hit Rate	Percentage	34.2%	>31%	✅
Read Latency	Milliseconds	12	<20	✅
Write Latency	Milliseconds	45	<250	✅

Performance metrics are automatically updated by CI/CD pipeline. Last update: see workflow runs

EOL is a comprehensive AI framework for building intelligent, context-aware applications.

Overview

This is a monorepo containing multiple packages for building intelligent AI applications with retrieval-augmented generation (RAG) capabilities.

Packages

Package	Status	Description	Test Coverage
eol-rag-context	✅ Active	RAG-based context management MCP server
eol-core	🔄 Planned	Core framework utilities	-
eol-cli	🔄 Planned	Command-line interface	-
eol-sdk	🔄 Planned	Python SDK for RAG applications	-

Key Features

• 🚀 Ultra-Fast Dependencies: Advanced wheel caching for 3-6x faster CI/CD builds
• 🔍 Intelligent Indexing: Content-aware chunking with AST parsing for code and semantic splitting for text
• 📊 Redis Vector Database: High-performance semantic search using Redis Stack v8
• 🧠 Semantic Caching: 31% cache hit rate target to reduce LLM API calls
• 🔗 Knowledge Graphs: Entity relationship mapping for enhanced context
• 📡 MCP Integration: Model Context Protocol server for seamless AI integration
• 👁️ File Watching: Auto-indexing with real-time change detection

Development

Requirements

• Python 3.13 or higher
• Redis Stack v8+ (with vector search module)
• UV package manager

Installation

This project uses UV for dependency management and workspace management.

# Install dependencies
uv sync --all-packages

# Run tests
./test_all.sh

License

GPL-3.0