Optimization Recommendations Built on AI Forecasts
Spacewise is a Zurich PropTech platform for the flexible leasing of pop-up shop spaces. On the platform's data, we trained forecasting models for individual locations and periods and derived optimization recommendations from them.
- Client
- Spacewise
- Status
- Completed
“What Voviva put together in such a short time really impressed me. Working with the team was easy and personal; you can tell they truly care.”
Chalid El AshkerCEO, SpacewiseAt a glance
A clean data foundation
Existing data flows through reproducible pipelines into training and forecasting.
Forecasts per pop-up location
The models predict how individual locations will develop over a given period.
Concrete optimization recommendations
The forecasts turn into an optimization recommendation for each location and period.
The challenge
Through Spacewise and its POP UP SHOPS marketplace, owners and operators lease spaces for pop-up shops flexibly and at short notice, for example in shopping centres, at stations, at markets or at events. Over the years, this has produced large amounts of data on the locations and how they have developed.
The data was there, but not yet usable for decisions about the pop-up spaces. What was needed was a foundation that turns the historical data into reliable forecasts and derives concrete optimization recommendations per location and period from them.
Our solution
We turned the data into clean, reproducible pipelines and trained scalable forecasting models on it that predict how a pop-up location will develop over a given period. With billions of data points, that called for memory-efficient processing and robust feature engineering.
The optimization recommendations build on the forecasts and provide a data-driven foundation for decisions about a location and period. Everything runs in Spacewise's Google Cloud environment, with Vertex AI for training and operation, from data ingestion to the recommendation.