Documentation

Getting Started Coming Soon

Conductor runs as part of the BespokeTracker stack. Set up the project hierarchy, configure your analysis engines, and start running simulations. Not yet publicly available -- check back soon.

Project Hierarchy

Conductor models work in a four-level hierarchy stored in the graph database:

  • Portfolio -- a collection of programs and standalone projects
  • Program -- a group of related projects with shared objectives
  • Project -- a bounded initiative with milestones and deliverables
  • WorkPackage -- an atomic unit of work with duration, cost, and resource assignments

Additional node types -- Milestone, Resource, Risk, Benefit, and StrategicObjective -- connect to this hierarchy through typed relationships. EVM metrics aggregate bottom-up from work packages through the full hierarchy.

Analysis Engines

Six engines operate on the same project graph:

  • CPM -- critical path, float, zero-slack identification
  • RCPSP -- resource-constrained scheduling via OR-Tools CP-SAT
  • EVM -- earned value metrics with four-level aggregation
  • Monte Carlo -- 10K-sample simulation with Bayesian learning
  • Portfolio Optimization -- mean-variance with graph-structural covariance
  • What-If -- in-memory scenario comparison against baseline

API Access

All analysis engines are accessible through the REST API at /api/conductor/. Key endpoints:

# Run critical path analysis
GET /api/conductor/projects/{id}/cpm

# Run Monte Carlo simulation
POST /api/conductor/projects/{id}/monte-carlo

# Get EVM metrics at any hierarchy level
GET /api/conductor/projects/{id}/evm

# Run portfolio optimization
POST /api/conductor/portfolios/{id}/optimize

# What-if scenario
POST /api/conductor/projects/{id}/what-if

Full API reference and advanced configuration are available in the GitHub repository .