Afya Intelligence is excited to announce its collaboration with Insupply Health and the Ministry of Health in Tanzania to develop Tabiri, AI predictive models for healthcare commodities quantification. This project can now start after completing a comprehensive review of existing methods like bottom-up quantification (BUQ) and the identification of several challenges that could be addressed by AI.
To ensure the success of this project, a workshop was held with the National Quantification Team (NQT) in Tanzania, aimed at reviewing the processes and tools for essential healthcare commodities quantification in Tanzania, highlighting areas for improving the current processes. Participants were able to share their experiences and identify the issues that have hindered the efficient and effective quantification of healthcare commodities in Tanzania.
“Tabiri models will be built to better represent the relation between different factors affecting the consumption of essential healthcare commodities”
During the workshop Afya Intelligence presented best practices in AI for healthcare commodities quantification, with a focus on successful case studies from other countries or regions and explored opportunities for AI in healthcare commodities quantification in Tanzania. This allowed participants to brainstorm and identify potential use cases for AI in healthcare commodities quantification, highlighting the potential benefits and challenges of implementing AI in Tanzania.
One of the identified challenges is the lack of technical capacity to deal with BUQ data and metrics, which can lead to errors and inaccurate quantification. Additionally, there is a need to improve forecasting accuracy, as the current method only considers the previous year's consumption without considering trends or regional differences.
To address these challenges, Afya Intelligence, inSupply Health and the Ministry of Health will be developing Tabiri, a predictive analytics solution that harnesses the power of AI to transform healthcare commodities quantification. Tabiri will provide an added advantage of annual trends, instead of just the previous year's consumption without considering whether there is a general upwards or downwards trend on the actual consumption.
Moreover, Tabiri models will be built to better represent the relation between different factors affecting the consumption of essential healthcare commodities, thus increasing consistency in quantification. It will also provide an overview dashboard that displays essential metrics for decision-making, such as order fill rates and products most likely to expire.
Use Case 1 : Anomaly Detection One of the key challenges in the quantification process is detecting outliers in the verification process. With Tabiri, we aim to develop an AI algorithm that can identify outliers in values at any level of quantification. This will help reduce the time used to assess quantification and ensure that errors are not missed.
Use Case 2: Improved Forecasting Accuracy Another challenge is improving forecasting accuracy. With Tabiri, we aim to provide an added advantage of annual trends instead of just the previous year's consumption without considering whether there is a general upwards or downwards trend on actual consumption. This will save time and provide consistency, ensuring that different facility workers factor in variables equally. Moreover, the algorithm would highlight similarities and differences in geographical locations, giving a bird's eye view of the quantification process.
Use Case 3: Predictive Analytics Dashboards Tabiri aims to provide overview dashboards that display all the metrics required for decision making, such as order fill rates, products most likely to expire, and forecasting accuracy. This will enable relevant stakeholders to make informed decisions and focus on important discussions during meetings. Moreover, the dashboard will highlight KPIs that need immediate attention, enabling prompt action. We believe that Tabiri will transform healthcare services and enable healthcare providers to make data-driven decisions. Stay tuned for more updates on Tabiri's development, and fill in the form below for more information or potential collaboration.
