Alfred is a sophisticated Perl-based AI assistant that leverages the Ollama API to provide intelligent responses with memory capabilities. Built with modern Perl features (5.38+), it offers a robust and extensible framework for AI interactions.

• Intelligent Memory System: Stores and retrieves relevant past interactions using Qdrant vector database
• Context-Aware Responses: Generates responses based on previous conversations
• Concept Extraction: Automatically identifies key concepts from prompts
• Configurable: Easy to customize through environment variables
• Modern Perl: Utilizes Perl 5.38+ features including experimental class syntax

• Perl 5.38 or higher
• Ollama server running locally or accessible via network
• Qdrant vector database (via Docker)
• Required Perl modules (see Installation)

1. Clone the repository:

git clone https://github.com/yourusername/alfred.git
cd alfred

2. Install dependencies:

cpan install LWP::UserAgent JSON HTTP::Request Dotenv

3. Start Qdrant using Docker:

docker run -p 6333:6333 \
    -v $(pwd)/qdrant_data:/qdrant/storage \
    qdrant/qdrant

4. Configure your environment:

cp .env.example .env
# Edit .env with your configuration

Create a .env file with the following settings:

OLLAMA_BASE_URL=http://localhost:11434
OLLAMA_DEFAULT_MODEL=llama2
OLLAMA_TEMPERATURE=0.7
OLLAMA_TOP_P=0.9
OLLAMA_MAX_TOKENS=4096

# Qdrant Configuration
QDRANT_HOST=http://localhost
QDRANT_PORT=6333
QDRANT_COLLECTION=memories
QDRANT_VECTOR_SIZE=384
QDRANT_DISTANCE=Cosine

Run Alfred with a prompt:

perl bin/alfred.pl "Your question here"

alfred/
├── bin/
│   └── alfred.pl          # Main executable
├── lib/
│   └── Alfred/
│       ├── Config.pm      # Configuration management
│       ├── HttpClient.pm  # HTTP client wrapper
│       ├── MemoryService.pm # Memory and concept management
│       ├── OllamaClient.pm # Ollama API client
│       └── Qdrant/
│           └── Client.pm  # Qdrant vector database client
├── .env                   # Environment configuration
└── Makefile.PL           # Build configuration

• Stores and retrieves conversation history using Qdrant
• Extracts key concepts from prompts
• Calculates similarity between prompts
• Provides context for new queries

• Interfaces with Qdrant vector database
• Manages vector storage and retrieval
• Handles similarity searches
• Configurable distance metrics

• Interfaces with Ollama API
• Handles text generation
• Manages model interactions
• Configurable generation parameters

• Manages HTTP requests
• Handles API communication
• Provides error handling
• Supports custom configurations

• Centralizes configuration management
• Loads environment variables
• Provides default values
• Validates settings

1. Install development dependencies:

cpanm --installdeps .

2. Run tests:

prove -l t/

1. Fork the repository
2. Create a feature branch
3. Commit your changes
4. Push to the branch
5. Create a Pull Request

This project is licensed under the MIT License - see the LICENSE file for details.

• Ollama for providing the AI backend
• Qdrant for vector database capabilities
• Perl community for modern language features
• Contributors and maintainers