Machine Learning-Based Forecasting for Value-Based Entities

About this project

For entities in the healthcare industry that bear financial risk, having the ability to predict spending is paramount. Our team was engaged to develop a predictive feature that allows for accurate forecasting of healthcare expenditures at both micro and macro levels. The challenge was not just in forecasting but in ensuring that predictions were continually refined and adapted to the evolving data landscape.

By analyzing historical claims data from various healthcare entities within a metroplex, combined with the scale of patients treated, our team devised a machine-learning algorithm. This powerful solution could forecast spending from 3 months up to 18 months, specific to entities like hospitals, nursing homes, or similar facilities, with an error rate of 5%.

Domain & Technology
Our Services Include


CHWE reached out to the Sobah team for this project after failing to get required delivery from 3 different partners in the past.

This problem involves data sources from over 50,000 different sources, where most of the data is unstructured.

Additionally, the user experience of the portal required encouraging citizens to be involved in their politics – a challenge that has eluded younger generations over the last few decades. Most importantly, the initiative is non-partisan and required completely neutral and data driven content to avoid any bias in the insights provided to the citizens.


The client required our team to conduct a thorough and careful study of data sources, architectural choices and innovative search engine optimization techniques. Product involved a sophisticated phased roadmap required to be delivered over a period of years. The core requirement was to build a highly scalable, adaptable and agile product framework and architecture. We were also tasked with engaging content writers to deliver thousands of articles for educating citizens on various topics related with their politics and society. Product envisioned additional roles and content created for analysts and institutions that require bulk data and deeper insights using data analytics and AI.

Our team

Upon integration, the algorithm not only produced monthly forecasts but also became progressively refined in accuracy with each new data input. This dynamic adaptability provided the healthcare entities with a reliable tool, ensuring they were better equipped to make informed financial decisions.

Got an idea?

Let's participate in the process of creation