demand-planning

Deterministic and Stochastic Dynamic Programs for optimization of Supply Chain

Forecast-driven inventory optimization project for retail demand planning, combining SARIMAX, ML model comparison, feasibility auditing, Monte Carlo simulation, and inventory policy optimization.

demand planning engine that combines probabilistic forecasting, conformal prediction, and ordering policies into a single backtestable pipeline

Creating Supply & Demand during tough times of lockdown caused by COVID-19

A spare engine placement generator based on a Finite-Horizon Markov Decision Process

Multi-model time-series forecasting with Bayesian Optimisation (Optuna TPE): SARIMA, Random Forest, XGBoost, LightGBM, Prophet, LSTM, and QuantileML probabilistic forecasts behind a unified ModelSpec protocol. Walk-forward validated; supports monthly, weekly, daily, and hourly data.

DuckDB extension for Git-like database branching. Create isolated scenarios for what-if analysis with copy-on-write storage, diff comparisons, and audit trails.

Zero-dependency demand forecasting for seasonal businesses. Pure statistics, no ML, no cloud costs.

o9 Solutions is an enterprise AI platform for integrated planning and decision-making, founded in 2009 by Sanjiv Sidhu (previously founder of i2 Technologies) and Chakri Gottemukkala, and headquartered in Dallas, Texas.

Executive-level B2B Sales Forecasting & Revenue Trend Analysis dashboard built with HTML, CSS, and JS, featuring a Pandas aggregation pipeline to analyze 100k+ transactional sales records.

Warehouse demand forecasting — Prophet, ARIMA & ensemble models. Stock alerts, What-if simulator, PDF purchase orders. 45 French SKUs, 11-tab dashboard.

Logility is an AI-powered supply chain planning platform providing solutions for demand sensing, inventory optimization, supply planning, S&OP process management, and supply chain analytics.

Demand shaping through pricing and promotion optimization

Sales forecasting model utilizing Armstrong Cycle Transformers to predict supermarket demand patterns.

Pharmacy demand forecasting that admits when naive wins: rolling-origin backtest across 5 model families with a beat-naive promotion gate (naive kept 13/40 SKUs), pooled-WAPE evaluation, Croston for intermittent demand, and an inventory simulation that prices forecast error as stockout-vs-waste cost.

Python demand forecast for pipeline throughput using real CER data — 30/60-day projection with a transparent moving-average model, explicit assumptions, and honest uncertainty ranges.

Demand regime change detection

Aggregate production planning linear programming

Is your MAPE lying to you? Five baseline models, rolling-origin backtesting, WMAPE/bias/FVA - and a Streamlit app to score your own demand data.

Demand-planning diagnostic toolkit: demand segmentation (ADI/CV2), forecast-accuracy scorecard, and Forecast Value Added on synthetic FMCG data. Stdlib-only, zero dependencies.