⚡ OptimEngine — Operations Intelligence Solver

⚡ OptimEngine — Operations Intelligence Solver

The first MCP Server for operations optimization across 4 intelligence levels.

Solves scheduling, routing, and packing — then quantifies risk, finds trade-offs, and prescribes actions. Built for AI agents in the agentic economy.

Intelligence Levels

Level 1 — Deterministic Optimization

Level 2 — Optimization under Uncertainty

Level 2.5 — Multi-objective Optimization

Level 3 — Prescriptive Intelligence

MCP Tools

Pricing

OptimEngine is free during beta. All 9 tools, all 4 intelligence levels.

Beta pricing ends when usage thresholds are reached. Get started free →

Quick Start

Install & Run

git clone https://github.com/MicheleCampi/optim-engine.git
cd optim-engine
pip install -r requirements.txt
uvicorn api.server:app --host 0.0.0.0 --port 8000

Connect via MCP

{
  "mcpServers": {
    "optim-engine": {
      "command": "mcp-proxy",
      "args": [
        "https://optim-engine-production.up.railway.app/mcp"
      ]
    }
  }
}

Examples

Scheduling (L1)

curl -X POST https://optim-engine-production.up.railway.app/optimize_schedule \
  -H "Content-Type: application/json" \
  -d '{
    "jobs": [
      {"job_id": "J1", "tasks": [
        {"task_id": "cut", "duration": 30, "eligible_machines": ["M1", "M2"]},
        {"task_id": "weld", "duration": 20, "eligible_machines": ["M2"]}
      ], "due_date": 80}
    ],
    "machines": [{"machine_id": "M1"}, {"machine_id": "M2"}],
    "objective": "minimize_makespan"
  }'

Prescriptive Intelligence (L3)

curl -X POST https://optim-engine-production.up.railway.app/prescriptive_advise \
  -H "Content-Type: application/json" \
  -d '{
    "solver_type": "scheduling",
    "solver_request": {
      "jobs": [{"job_id": "J1", "tasks": [
        {"task_id": "cut", "duration": 30, "eligible_machines": ["M1"]}
      ], "due_date": 80}],
      "machines": [{"machine_id": "M1"}],
      "objective": "minimize_makespan"
    },
    "forecast_parameters": [{
      "parameter_path": "jobs[J1].tasks[cut].duration",
      "historical_data": [
        {"period": 0, "value": 25}, {"period": 1, "value": 28},
        {"period": 2, "value": 30}, {"period": 3, "value": 33},
        {"period": 4, "value": 35}
      ],
      "forecast_method": "exponential_smoothing"
    }],
    "risk_appetite": "moderate"
  }'

Use Cases

• Manufacturing: Production scheduling with demand forecasting and risk quantification
• Logistics: Delivery routing with variable demand, travel times, and fleet trade-offs
• Warehouse: Bin packing with uncertain item weights and multi-objective optimization
• Supply Chain: End-to-end prescriptive intelligence: forecast → optimize → advise
• Finance: Portfolio-like resource allocation under uncertainty with CVaR metrics

Architecture

AI Agent (Claude, GPT, Gemini, etc.)
    │
    ▼ MCP Protocol
┌──────────────────────────────────────────────────────┐
│  FastAPI + fastapi-mcp                                │
├──────────┬──────────┬──────────┬─────────────────────┤
│  L1      │  L1      │  L1      │  L2  L2.5  L3       │
│ Schedule │ Routing  │ Packing  │ Sensitivity          │
│ CP-SAT   │ Routing  │ CP-SAT   │ Robust               │
│          │ Library  │          │ Stochastic            │
│          │          │          │ Pareto                │
│          │          │          │ Prescriptive          │
├──────────┴──────────┴──────────┴─────────────────────┤
│  OR-Tools Solvers + Monte Carlo + Forecasting Engine  │
├──────────────────────────────────────────────────────┤
│  Pydantic v2 Models                                   │
└──────────────────────────────────────────────────────┘

Tests

python -m pytest tests/ -v

121 tests across 9 modules.

Landing Page

🌐 optim-engine-landing.vercel.app

Marketplace Listings

• MCPize — 11 MCP tools
• Apify Store
• LobeHub
• mcp.so

License

MIT

Built by Michele Campi — Operations Intelligence EngineerThe first MCP server with 4 levels of optimization intelligence.