Livestock Data Measurements & Analysis

Barefoot Lightning's products target underserved livestock farmers in developing countries filling major gaps in the availability of effective veterinary services, excessive feed costs which contribute to poor productivity and finally market price issues in the predominantly informal markets.

We use animation content to bridge the gap to low literacy farmers and have developed a significant scientifically based research database to drive decisions based on best available data. This still requires us to develop a number of data science models to deliver decision support to vets, farmers and extension workers.

We have worked to push the thinking as deep as we can with the access we've had to various scientists, but now that the base platform is built and our understanding of the challenges are clearer, we need World Class data science support to help us crack these challenges in the most effective ways possible, as well as to develop an AI based managed learning infrastructure so we can drive continuous improvement as well as human-based learning from the system.

The first major challenge we have is related to the Bayesian probability model we have developed for the symptomatic disease diagnostic tool which supports vets to more rapidly access case data and come to effective diagnoses for a much larger group of farmers, as well as supporting the delivery of these services through field trained individuals such as paravets and animal health care workers.

The next challenge relates to multivariate analysis to identify the minimum costs for a nutritionally balanced feed recipe based on the cost of locally available feed ingredients. We hope to improve the function of the current linear simplex-based model and also build in an AI based learning mechanism to identify ingredients' impacts beyond their basic nutritional supply.