Forecasting for staff scheduling software
Lokad has been designed to be integrated into 3rd party applications. Our ambition is not to become the next killer inventory or staff management application, but to stick to what we do best: statistical forecasting.
Thus, I have just published a case study Forecasting for scheduling software that involves a fictitious company named Scheduloo delivering a staff scheduling software in ASP.
Disclaimer: All characters depicted herein are fictional characters, and any similarities to actual persons and/or companies are purely coincidental.
In summary, the case study details how (and why) Lokad could be integrated as a forecasting component in a staff scheduling application in order to provide a staff optimization feature.
The case study has been kept simple and tractable, A priori, it should not contain any rocket science ideas. Yet, more ambitious approaches could also be considered.
For example, following the idea of the evidence-based scheduling that ships with Fogbuzz, the mapping sales2staff (i.e. converting turnover into staff level) could be inferred from the historical data as opposed to be made explicit by the staff manager.
Then, many businesses (such as retail) are very dependent from the weather. Lokad, although being accurate for business forecasts, will only perform poorly on weather forecasts. Thus, there is a nice mash-up to be made that would combine the Lokad technology with classical weather forecasts as provided by the National Weather Service for example.
