Projects
Modeling the Mortality and Morbidity Risk of COVID-19 and Influenza due to Demographic, Socio-Economic, and Behavioral Factors in the USA
We used county-level demographic, socio-economic, and behavioral data to construct risk factors associated with population, age, household size, income, mobility, and mask usage and constructed a risk score for each county in the USA. Risk scores were computed for cases and deaths and hence morbidity and mortality. Principal components were constructed to eliminate linear dependence of the predictors, and polynomial regression was used to arrive at the risk-scoring equations. The process was invertible to explain the leasing cause of high risk in a particular county. The process was repeated for Influenza to compute state-level risk scores for each year since 2013. The results were provided to government agencies and health systems to monitor the spread of the epidemic and plan vaccination drives. The information was also published in a portal for public awareness.
Data Warehousing and Analytics for Public Health Insurance
The project is about providing a next generation and state-of-the-art Analytical Ecosystem as required at Dr.NTRVST for addressing the two fold objective of providing descriptive and predictive analytics to the stakeholders at Dr.NTRVST. The proposed Analytical system should help users move from the existing operational reports available in the Dr.NTRVST Transactional system that are static in nature and create a Self-Service Analytical Environment where users can slice and dice the Dr.NTRVST Data and gain insights into the data by applying various Statistical and Data Mining techniques. The Analytical system should seek to automate various activities that are currently done manually while trying to perform analysis on the Dr.NTRVST data and also bring-in the best practices to the Analytical System. The end-result of identifying exceptions and hidden trends in the Dr.NTRVST data should be achieved with minimal manual interventions and the overall decision making and planning process needs to be made more efficient.
Products
Scalable Platform to Provide Actionable Information for Health Insurance Decisions
We built a product that effectively integrates clinical, claims, psychographic, and physiologic data from a variety of sources into one platform and presents a holistic patient view through an intuitive dashboard that supports physician decision-making and care team action, improving financial, quality, and health outcomes. The product included a highly effective clinical data model and predictive algorithms that exceed the industry norms for predictive accuracy. Traditionally, health plans have used actuarial models that determine risk and future costs at the population level. Our platform used leading-edge machine learning techniques to create a dynamic risk score for each member, which improves its precision and personalization as well as enables health plans to act on the insights to prevent health deterioration and future costs.
Letter of Protection Management System
We created a web-based medical practice management system for linking patients to doctors and attorneys, scheduling medical procedures, and tracking the resulting claims. The Application is optimized for personal injury business processes. It provides a robust system for tracking claims throughout their lifetime which includes receiving new orders, requesting/receiving documents between attorneys and physicians, scheduling patients for their procedures, and generating reports and invoices. The application is designed in a multi-tenant architecture so that it can onboard different companies onto it’s system and process their requests concurrently. The system’s Document system is also designed to be HIPAA compliant for better security. The backend of the application is deployed in Google Cloud Platform using cloud Datastore and cloud storage as its main Database,while it is powered by a Flask server deployed in Google Compute Engine. The frontend of the Application is designed in Vuetify which is hosted in google firebase hosting.
Consulting
Freight Assessment for Regional Waterway Connectivity
The technical, socio-economic, and logistical aspects of constructing a new canal to establish a direct inland connection between the Hooghly/Ganga River in India and the Padma River in Bangladesh, led to modeling the impact of the proposed canal on the pattern of transportation. The canal would not only be effective to divert a significant amount of traffic from the current land routes used for India-Bangladesh trade but could potentially divert trade from the North-East region to the rest of India. Data was collected from various government sources and trade-route partitioning was modeled based on the origin and destination of goods and the border crossing chosen for each goods category. The result was then mapped to a new potential waterway channel and the potential transition was estimated. This result was then forecast for the next thirty years.
Intra-City Freight Flow in the City of Kolkata
Fright transport around the city of Kolkata is complicated by the presence of River Ganges which requires traversal of the river across two main bridges resulting in congestion and delays. The World Bank Group intended to build roll-on/roll-off jetties at several locations to augment the available routes for trucks to cross the river. A study was commissioned to identify the impact of intra-city freight movement in the city of Kolkata as a result of these new crossings. The city was geographically divided into natural regions demarcated by canals or railway lines such that there were limited ways to cross from one region to the other. At each crossing, surveyors counted all goods transport vehicles while also recording their size and the type of goods carried. The survey lasted one week at each location, twenty-four hours a day. An additional survey was conducted at each large market in the city to interview shop owners to gauge and understand delivery patterns. The results were combined into a topological model which resulted in a visualization of the intra-city traffic flow by type of goods and vehicles across time and day of the week. As a result, the model could deliver a view of the potential consequences of local disruptions and deviations and was used to plan the location of the proposed jetties.
Research
Scalable Platform to Provide Actionable Information for Health Insurance Decisions
We built a product that effectively integrates clinical, claims, psychographic, and physiologic data from a variety of sources into one platform and presents a holistic patient view through an intuitive dashboard that supports physician decision-making and care team action, improving financial, quality, and health outcomes. The product included a highly effective clinical data model and predictive algorithms that exceed the industry norms for predictive accuracy. Traditionally, health plans have used actuarial models that determine risk and future costs at the population level. Our platform used leading-edge machine learning techniques to create a dynamic risk score for each member, which improves its precision and personalization as well as enables health plans to act on the insights to prevent health deterioration and future costs.