dcisionai-mcp-server 1.4.0

DcisionAI MCP Server for Mathematical Optimization with Enhanced Solver Selection and Business Explainability

DcisionAI MCP Server

A Model Context Protocol (MCP) server that provides real mathematical optimization capabilities powered by OR-Tools and AI-driven problem formulation.

Features

• 6 Core Optimization Tools: Intent classification, data analysis, model building, optimization solving, workflow templates, and end-to-end execution
• Real Mathematical Optimization: Uses OR-Tools for genuine mathematical solving (not AI generation)
• AI-Driven Formulation: Leverages Qwen 30B for intelligent problem formulation
• Industry Workflows: Pre-built templates for manufacturing, healthcare, retail, marketing, financial, logistics, and energy sectors
• MCP Protocol: Seamless integration with AI development environments like Cursor

Installation

pip install dcisionai-mcp-server

Usage

The server can be run directly:

dcisionai-mcp-server

Or via uvx:

uvx dcisionai-mcp-server

Configuration

Set the following environment variables:

• AWS_ACCESS_KEY_ID: Your AWS access key for Bedrock access
• AWS_SECRET_ACCESS_KEY: Your AWS secret key for Bedrock access
• AWS_REGION: AWS region (default: us-east-1)

Tools

1. classify_intent: Classify user intent for optimization requests
2. analyze_data: Analyze and preprocess data for optimization
3. build_model: Build mathematical optimization model using Qwen 30B
4. solve_optimization: Solve the optimization problem using OR-Tools
5. get_workflow_templates: Get available industry workflow templates
6. execute_workflow: Execute a complete optimization workflow

License

MIT License