Demand & Capacity Planning Platform with Forecasting Intelligence

Service departments in municipalities can find themselves in a reactive staffing situation and are often over or under staffed for a time, which has a negative effect on service quality and a negative impact on the operational costs. The project meets the requirement of Ajman Municipality to come up with a data-driven solution for demand forecasting and workforce planning.

 

CapaCast is an AI-driven demand and capacity planning platform created in partnership with the Data Analytics Program from Ajman University. Historical transaction data is fed into the system from four municipal services: Commercial Contracts, Contract Cancellation, Register Property and Residential Contracts; and Gradient Boosting Regression is used along with engineered lag and rolling-window features to produce weekly and monthly demand forecasts. A dashboard in English/Arabic with interactive charts, peak/low demand indicators and staffing recommendations using a 77requests per day capacity sensitivity is used to display predictions that are generated by a FastAPI backend. This is a platform that integrates a Gmail alert mechanism for when demand is anticipated to surpass capacity with Ollama LLM for creating business reports, Alert Gmail Structure and Report generation via local AI. CapaCast provides measurable value from a business perspective by optimizing staffing, managing resources in advance, and cutting back on operating expenses with an accuracy of 87%–97% for all services. A new Business Success module monitors KPIs, cost benefit results and stakeholder feedback to validate the effect of the platform. CapaCast allows municipality planners to move from reactive to pro-active decision making and has a scalable architecture that is suited to integrate live SQL and expand across multiple departments.