During COVID, I noticed that many people seeking medical care needed help finding the right provider (physician) that fit their needs. It was always difficult to find the availability of providers online and how an application can help patients to search for providers and book appointments whenever they need for their health concerns. For this, I designed a Provider Search web application that allows people to search their providers based on their needs and returns a series of appropriate providers in a minuscule amount of time.

What it does

Today, the healthcare system is very complex and we want to make sure that the patient should get high-quality care in a cost-effective manner. A provider search web application is a tool with intelligent search application that is built on a knowledge base. It understands the patient's needs and provides a list of ranked providers that are most applicable to the patient factors. The list of providers returned from the search are factors based on the patient’s insurance, providers’ in/out network relationship, patient’s health concern, patient’s location, etc. Each of these factors helps to boot the provider and calculate the appropriate score to get providers in the right order when they are presented to the user.

How I built it

As we know, it is a very complex process to surface the right set of providers that matches the patient's needs. So, I decided to use: -Node js which is a javascript runtime server environment that serves as a web server for user interface capabilities. -REST APIs communicate JSON payload back and forth from the user interaction to the server.

  • The provider knowledge base is hosted on the OpenSearch search engine.
  • The boosting algorithm is used in the OpenSearch query to get the most appropriate providers at the top of the list. ## Challenges I ran into It is very important to calculate the document score based on the provider matches along with document filtering. It was extremely challenging for me to identify the right set of OpenSearch query constructs that have boosting logic and return documents matching a positive query while reducing the relevance score of documents that also match a negative query along with filtering records. ## Accomplishments that we're proud of I am proud that I was able to accomplish the search matching algorithm built as a result of the OpenSearch query to get optimal out-of-the-box performance. Thanks to the OpenSearch specification offering compound queries like function_score, script_score, geo_distance, etc. ## What we learned This project helped me to learn in various areas that include
    • Understanding how to build complex web pages using handlebars and javascript.
    • Understanding and building REST-based APIs
    • Understanding OpenSearch Search Engine capabilities and functionalities. ## What's next for Provider Search The application has been built for the hackathon purpose and truly I would like to extend the application to support online appointment scheduling capabilities based on provider availability, adding more boosting capabilities on licensing providers, filtering by age limit, etc.
Share this project: