Pareto-Driven Bottleneck Detection from Operational Data
N.B. Upload comp history (salary, bonus, grant-date value, realizable value), metric results, peer list, and share price history. Confirm whether to use grant-date or realizable pay and whether TSR is relative or absolute. Ensure currency conversion rules and FX rates are consistent across years.
📄 Prompt Template
Analyze historical data from [DataSources] across [RoutesOrLanes] for [PeakPeriod] to locate the operational bottleneck impacting [Metric] (e.g., on-time departure, dwell, throughput). Create a Pareto view of delay and variance contributors by node (port/terminal/cross-dock/warehouse), activity, and cause code. Segment by [SKUGroup] to reveal mix effects.
Output format:
Insight Summary (≤170 words).
Markdown table “Top 10 Loss Buckets” with columns: Node, Activity, Cause, Frequency, Total Hours Lost, % of Total, Cumulative %.
Bottleneck Candidate profile (one-pager) detailing process map, average CT, p95 CT, queue length proxy, rework rate, and data coverage %.
Three testable root-cause hypotheses with required diagnostics and quick-win countermeasures.
Control chart specification (metric, sampling plan, UCL/LCL approach) for post-fix monitoring.
Ensure reproducibility: include source fields and join keys.