The client is a provider of online tax preparation solutions.
The Challenge
Our client offers online tax preparation solutions, allowing American consumers and tax professionals to electronically file state and/or federal returns. While this definitely makes life easier for taxpayers, it also presents multiple challenges for the firm. Tax preparation is, by nature, a cyclical business, driven by the annual April filing deadline, as customer queries spike during the peak tax preparation period, leading to scalability issues. Each year, the firm hires customer service contractors to keep up with surging demand, which leads to knowledge-sharing difficulties – new hires face a steep learning curve, with only limited knowledge repositories and few in-house experts to turn to for assistance. Predictably, response times for customer queries suffer. The CTO and CIO agreed that deploying an AI-powered support solution would improve their customers’ experience, reduce costs, and help drive company growth.
The Solution
Persistent crafted a three-phase solution to answer immediate challenges while putting the client on a solid long-term footing. We began by setting up a Central Knowledge Repository to help representatives answer customer queries, integrating information from multiple repositories that were previously separate and siloed. Next, we developed a Conversational AI Teams BOT – an intuitive interface for obtaining summarized answers to complex customer queries. Finally, we added an Automated Data Pipeline, i.e., a real-time knowledge base feed to ensure that responses to customer queries were accurate and up to date.
We leveraged Microsoft Cloud, AI, and Generative AI technologies throughout the solution build, including Azure Stack for scalability and security, Azure Open AI Models for knowledge capture and integration, and Azure Cognitive Search and BOT Service for the application front-end. The integration of Azure cloud and GenAI capabilities resulted in a faster and less costly development cycle, as well as a better end result.
The Outcome
The Persistent solution delivered ~80% faster turnaround times in resolving customer queries, plus, more importantly, ~75% improvement in response accuracy. To date, the solution has yielded a ~40% savings in total FTE cost while improving the customers’ experience and supporting the larger goal of continued company growth.
Technology used:
- Azure Stack
- Azure Open AI Models
- Azure Cognitive Search
- Azure BOT Service
