One Benefit solution

What Inspired us

In this current pandemic situation, We have seen customer queries regarding their health plan coverages and benefits becoming complex and Customer Service representatives trying their level best to toggle between diff systems trying to accurately answer those questions , we wanted to provide a solution which would help the customer representative look for required information in one place , allowing the CSR to provide accurate information as possible and also manage to cut down the Average Handling time on the call ,resulting in better customer service . Benefit eligibility query processing in HC payer contact centers takes long (avg ~30 mins) as agents need to go through multiple disparate systems and documents, often resulting in unsatisfactory, inadequate response to queries, resulting in:

Large Operational Overhead
Inefficient manual interaction reduces agent productivity
Lack of intelligent query lookup results in High AHT & Reduced FCR
Wrong results bring in costly follow-up like Authorization Member / Provider Abrasion
High TAT and suboptimal customer experience
Revenue loss from churn
Lack of Accuracy
Member needs to provide specifics search options
Lack of confidence level/ accuracy factor for the result

How we built it

As part of our solution , we would be providing the CSR with an option to search with generic keywords on diff procedures and common industry terminologies used , which in turn will be hitting a NLP Processing algorithm ( Natural language processing Algorithm) to find a match in the CPT and ICD code database and rank the search results as per relevance and display it o the CSR , so the ycan answer customer queries. Benefit SOR/ grid based determination is used to identify the appropriate benefit details. NLP (Natural Language Processing ) along with Self learning AI based solution will provide 95+% accuracy in retrieving Benefit information for a customer .

This Pega based decoupled solution leverages below solution components as the customer service agent enters the member query in common English:

Medical / Clinical Search It leverages NLP processing, Checks medical ontology and identifies match in the form of CPT / ICD code.
Benefit Lookup Based on agent’s selection, it leverages Pega based rule engine to check Benefit System of Record /Benefit grid spreadsheets to identify the benefit
Benefit Analysis Leverages NLP processing to analyze indexed docs like EOB, EOC, cost sharing and results detailed result

Challenges we ran into

Trying to identify all possible generic questions asked by customers today with regards to their benefits and trying to build a self learning AI based on customer responses was a critical and challenging task. Addresses common problem across industries that calls for an channel & technology agnostic intelligent self-learning platform that can transform complex data into clear & instantaneous answers

Accomplishments that we're proud of

we were trying to find best repository which can translate simple 'english' term to more medical terminology e.g weight loss surgery to Bariatric surgery. we have tried multiple available API but it was not giving must match, near match. the terminology dictionary, Pega API and NLP engine with elastic search for document indexing has fulfilled all in-vision requirement. it's seamless one screen view which currently CSR might be spending multiple minute to shuffle between applications and documents

What we learned

We have learn secure API integration, pega robust NLP processing with feedback loop and elastic search capability. it's unique learning and presents how multiple technology and powerful pega technology can translate complex business problem at extremely simple view to end users.

What's next for HC- One Benefit solution

this idea can be translated to insurance, banking and other area. Same use case can be applied to other domain with different flavor of purpose like insurance coverage search, etc.