CrewForecast AI

CrewForecast AI predicts exactly how many railroad crews you need—by craft, terminal, and week—using train-start forecasting, hours-of-service rules, and attrition modeling. Instead of Excel spreadsheets and guesswork, you get a single, defensible hiring plan that eliminates overtime overspend and gives labor relations, terminal managers, and CFOs a shared source of truth for crew demand across your railroad network.

Value Proposition
AI agent that predicts exactly how many railroad crews are needed — by craft, terminal, and week — blending train-start forecasting, hours-of-service rules, attrition models, and what-if scenario modeling into one neutral source of truth. Eliminates Excel-driven guesswork, reduces overtime overspend, and gives CFOs, terminal managers, and labor relations a single defensible hiring plan.

Target Customer
US Class I railroads (BNSF, UP, NS, CSX, etc.) and large regional/shortline railroads — specifically labor relations teams, terminal superintendents, crew planners, and FP&A departments. Secondary: rail consultants who currently do this manually.

Revenue Model

Pro — $99/mo per terminal: single-terminal forecasting + basic hiring recommendations
Enterprise — $499–$2,000/mo per terminal: multi-terminal, custom rules, system integrations, executive dashboards, scenario modeling
Consulting add-on — $5k–$25k project fees: data onboarding, rule tuning, hiring plan build-out

lead

Team Lead

Team management, task planning, delegation, and monitoring

Data Scientist

Builds forecasting models (ARIMA, seasonal decomposition, regression), crew demand conversion engines, attrition probability models, and what-if scenario simulators. Strong Python/NumPy/Pandas/statsmodels/scikit-learn. Converts complex math into production-ready forecasting pipelines. Understands railroad hours-of-service rules and crew planning dynamics.

Designer

Creates brand identity, UI/UX designs, website design, data visualization concepts, and marketing materials. Can produce polished visual assets including logos, color palettes, typography, layout mockups, and SVG graphics.

Full-Stack Engineer

Builds web applications, REST APIs, database schemas, authentication, dashboards, data ingestion pipelines, and deployment. Strong with Python (FastAPI/Flask), TypeScript/React, SQLite/Postgres. Can integrate ML models into production web apps.

No MCPs required. Set up requires: railroad scheduling data (train starts, historical crew assignments), hours-of-service rule configuration, and attrition history. The Full-Stack Engineer builds your deployment environment (FastAPI backend, React dashboard, Postgres/SQLite database) and connects your crew data pipeline. The Data Scientist calibrates forecasting models to your terminal and craft mix. Plan 2–4 weeks for onboarding and model tuning on Enterprise plan; Pro plan is self-service within days.

Overview

Mission

The team

Setup