🤖 AI Summary
Large-scale manufacturers face challenges in after-sales demand forecasting, including difficulty in fusing heterogeneous multi-source signals, weak modeling of COVID-19 disruptions, imbalanced prediction accuracy for long-tail versus high-revenue items, and insufficient business interpretability. Method: We propose an end-to-end interpretable ensemble forecasting framework integrating statistical models, deep learning, and large language models (LLMs). Key innovations include Pareto-aware segmented forecasting, horizon-aware weighted ensemble integration, LLM-driven automated attribution narrative generation, plus integrated change-point detection and WMAPE-optimized calibration. Contribution/Results: The system delivers city-item-level calibrated probabilistic forecasts across 90+ countries and 6,000 SKUs, simultaneously improving prediction accuracy, stability, and operational alignment. Through a performance scorecard and trend attribution module, it shifts evaluation from static accuracy metrics to a dynamic, intervention-oriented decision loop.
📝 Abstract
This paper presents a practical architecture for after-sales demand forecasting and monitoring that unifies a revenue- and cluster-aware ensemble of statistical, machine-learning, and deep-learning models with a role-driven analytics layer for scorecards and trend diagnostics. The framework ingests exogenous signals (installed base, pricing, macro indicators, life cycle, seasonality) and treats COVID-19 as a distinct regime, producing country-part forecasts with calibrated intervals. A Pareto-aware segmentation forecasts high-revenue items individually and pools the long tail via clusters, while horizon-aware ensembling aligns weights with business-relevant losses (e.g., WMAPE). Beyond forecasts, a performance scorecard delivers decision-focused insights: accuracy within tolerance thresholds by revenue share and count, bias decomposition (over- vs under-forecast), geographic and product-family hotspots, and ranked root causes tied to high-impact part-country pairs. A trend module tracks trajectories of MAPE/WMAPE and bias across recent months, flags entities that are improving or deteriorating, detects change points aligned with known regimes, and attributes movements to lifecycle and seasonal factors. LLMs are embedded in the analytics layer to generate role-aware narratives and enforce reporting contracts. They standardize business definitions, automate quality checks and reconciliations, and translate quantitative results into concise, explainable summaries for planners and executives. The system exposes a reproducible workflow -- request specification, model execution, database-backed artifacts, and AI-generated narratives -- so planners can move from"How accurate are we now?"to"Where is accuracy heading and which levers should we pull?", closing the loop between forecasting, monitoring, and inventory decisions across more than 90 countries and about 6,000 parts.