With the majority of East Africa being faced with severe market price shocks and food insecurity, the impact of Covid-19 has created further devastating impacts upon those already most vulnerable. Through Sauti Collect users are able to gain access to food prices in real-time to decrease the risk of price shocks and maximise economic productivity. 

This assignment calls for an automated market information system (MIS), a pilot of market information mapping and sharing via mobile, and piloting of a market intelligence and forecasting system. This will have the capacity to capture and present recent daily Kenyan market prices for various food products. 

Price information will be inputted by price providers who have been trained on how to submit prices of different commodities as well as the mapping of supply chains through USSD in the three targeted project markets. 

The project has piloted supply chain mapping in one target market, which has been selected as Nairobi. In order to map suppliers, Sauti designed data collection tools to be used with vendors in Nairobi. Data collection tools have been able to provide a deeper understanding of the market systems that vendors are working within, and their suppliers in select value chains.

The initial forecasting methodology will be designed by leveraging Big Data from additional market information price providers and deploying a high-dimensional adaptive quantitative analysis (e.g. elastic net regressions) in order to ensure that the forecasting outputs are resilient to market price data availability and forecasting challenges, and provide policy-relevant short-term forecasts. We will test the forecasting models with the market price data currently being collected by vendors and enumerators. This test phase will provide a current assessment of the model’s performance and highlight issues related to data cleaning, availability, and context.