OptimEngine — Operations Intelligence Solver (L1→L3)
11-tool MCP server for scheduling, routing, bin packing with sensitivity, robust, stochastic, Pareto & prescriptive intelligence. v9: per-machine speeds, quality gates, shift windows, setup matrix.
OptimEngine v9.0.0 — the first MCP Server for operations optimization under uncertainty.
11 tools across 4 intelligence levels: • L1 Deterministic: Scheduling (FJSP), Routing (CVRPTW), Bin Packing — all with 4 objectives • L2 Uncertainty: Sensitivity analysis, robust optimization, stochastic Monte Carlo + CVaR • L2.5 Multi-objective: Pareto frontier for 2-4 competing objectives • L3 Prescriptive: Forecast → optimize → risk assess → recommend
v9.0.0 scheduling upgrades: • Per-machine durations (CNC-1: 120min, CNC-2: 80min for the same task) • Multiple availability windows (shifts, maintenance, breaks) • Quality gates (job requires yield >97%, machine has yield 99%) • Sequence-dependent setup times (changeover matrix between product families)
Live-tested on MetalPrecision S.r.l. scenario: 5 automotive clients (Bosch, Ducati, Maserati, Comer Industries, Dallara), 4 CNC centers, 15 tasks — optimal in 0.04s, all 7 intelligence tools producing actionable results.
Built with OR-Tools CP-SAT, Python 3.12, FastMCP. 121 tests passing. Zero capital.