Consultant Utilization Forecasting and Capacity Balancing Engine

Role: Principal Workforce Intelligence Data Scientist for enterprise advisory firms.

Context

• Target practice divisions: {{firm_service_lines}}
• Staff allocation dataset: {{capacity_allocation_data}}
• Overtime warning ceiling: {{overtime_threshold_hours}}
• Forward projection window: {{forecast_horizon_weeks}}
• Required statistical confidence: {{statistical_confidence_level}}
• Webhook destination: {{alert_webhook_target}}

Task

Develop a modular Python or R predictive analytics script that forecasts consultant utilization across {{firm_service_lines}}, identifies impending bench surplus or burnout bottlenecks, and publishes payload summaries to {{alert_webhook_target}}.

Method

1. Ingest historical time tracking and forward pipeline allocations from {{capacity_allocation_data}} across the specified {{firm_service_lines}}.
2. Cleanse and normalize schedule records, accounting for statutory holidays, planned annual leave, and training buffers.
3. Fit a rolling time-series forecasting model projecting weekly consultant billable hours across {{forecast_horizon_weeks}}.
4. Calculate statistical forecast intervals at the {{statistical_confidence_level}} level to quantify staffing volatility.
5. Detect individual and practice-wide anomalies exceeding {{overtime_threshold_hours}} or dropping below baseline billable expectations.
6. Compute supply-demand gaps categorized by consultant seniority, skill tags, and location hubs.
7. Generate structured alert payloads containing high-risk capacity breaches and format them for {{alert_webhook_target}}.
8. Package the entire sequence into a deterministic batch job with comprehensive run diagnostics.

Constraints

• MUST set deterministic random seeds for all statistical estimators to guarantee reproducible forecasts.
• MUST NOT drop incomplete records silently; impute missing schedules using explicit, documented fallback strategies.
• The script must complete end-to-end execution in under 90 seconds for a dataset of up to 10,000 consultants.
• Payload formatting must strictly validate against standard webhook JSON specifications.

Output format

• Section 1: Forecasting methodology and assumptions matrix (100-150 words)
• Section 2: Complete production script with data processing, forecasting logic, and dispatch routine
• Section 3: Webhook payload schema documentation and sample output JSON (100-150 words)

Self-review

• Confirm that {{forecast_horizon_weeks}} and {{overtime_threshold_hours}} are correctly factored into threshold flags.
• Verify that webhook dispatch handles network timeouts and retries cleanly.
• Check that confidence interval calculations align with {{statistical_confidence_level}}.