Human-in-the-Loop Available-to-Promise Intelligence for Reliable Delivery Date Prediction

Available-to-promise (ATP) systems increasingly rely on machine learning to predict delivery dates under conditions of fluctuating supply, capacity constraints, and demand variability. Yet purely algorithmic ATP pipelines face a persistent adoption barrier: planners and account teams frequently override or distrust automated delivery-date predictions, particularly after observing even isolated forecasting errors, a pattern long documented in the algorithm-aversion literature. Conversely, human-only override regimes reintroduce the inconsistency and latency that automation was meant to eliminate. This paper proposes a human-in-the-loop (HITL) available-to-promise intelligence framework that resolves this tension by design rather than by policy. Following a design science research approach, the framework separates delivery-date prediction into four architectural layers: an algorithmic prediction layer, a confidence and uncertainty estimation layer, a human-review trigger layer governed by explicit escalation criteria, and an override/feedback-capture layer that closes the loop back into model governance. The trigger layer operationalises when a prediction should reach a human reviewer, using confidence thresholds, anomalous order-pattern detection, and value/risk tiering rather than blanket human review or blanket automation. The paper situates this design against literature on algorithm aversion and appreciation, human-in-the-loop machine learning architectures, and explainability in supply-chain artificial intelligence and evaluates the framework's utility against design science research criteria rather than primary empirical validation. The paper closes with an explicit limitations discussion and a concrete agenda for the multi-case empirical validation the framework's real-world deployment would require.

Authors

Publication Details

Journal
East African Journal of Information Technology
Published
2026-11-18
DOI
https://doi.org/10.37284/eajit.9.3.6324
Primary Topic
Impact of AI and Big Data on Business and Society
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Human-in-the-Loop Available-to-Promise Intelligence for Reliable Delivery Date Prediction

Venu Gopal Kakarla
3 links
East African Journal of Information Technology
Impact of AI and Big Data on Business and Society
article

Human-in-the-Loop Available-to-Promise Intelligence for Reliable Delivery Date Prediction

Venu Gopal Kakarla
article en

Abstract

Available-to-promise (ATP) systems increasingly rely on machine learning to predict delivery dates under conditions of fluctuating supply, capacity constraints, and demand variability. Yet purely algorithmic ATP pipelines face a persistent adoption barrier: planners and account teams frequently override or distrust automated delivery-date predictions, particularly after observing even isolated forecasting errors, a pattern long documented in the algorithm-aversion literature. Conversely, human-only override regimes reintroduce the inconsistency and latency that automation was meant to eliminate. This paper proposes a human-in-the-loop (HITL) available-to-promise intelligence framework that resolves this tension by design rather than by policy. Following a design science research approach, the framework separates delivery-date prediction into four architectural layers: an algorithmic prediction layer, a confidence and uncertainty estimation layer, a human-review trigger layer governed by explicit escalation criteria, and an override/feedback-capture layer that closes the loop back into model governance. The trigger layer operationalises when a prediction should reach a human reviewer, using confidence thresholds, anomalous order-pattern detection, and value/risk tiering rather than blanket human review or blanket automation. The paper situates this design against literature on algorithm aversion and appreciation, human-in-the-loop machine learning architectures, and explainability in supply-chain artificial intelligence and evaluates the framework's utility against design science research criteria rather than primary empirical validation. The paper closes with an explicit limitations discussion and a concrete agenda for the multi-case empirical validation the framework's real-world deployment would require.

East African Journal of Information TechnologyVol. 9(3)
Impact of AI and Big Data on Business and Society
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon