Client background

Baker Tilly is leveraging generative AI to develop a comprehensive knowledge management tool for working with Sage Intacct. This collaborative effort between the Sage Intacct Product Development and Advanced Business Technology Services teams aims to provide users with easy access to a wide range of information, including general Sage Intacct capabilities, available features, technical details for developers and functionalities specific to Baker Tilly's Sage Intacct product offerings. Functioning as an automated Subject Matter Expert (SME), the tool will serve internal users by streamlining access to knowledge on Sage Intacct and Baker Tilly's related products. Additionally, it will empower personnel to deliver superior client service and facilitate on-demand learning opportunities.

The business challenge

The leadership team wanted to create a generative AI application that would leverage a large repository of proprietary knowledge and documentation to generate context-aware responses so that their customer service, customer success and professional services teams could accelerate time to resolution and increase customer satisfaction. They approached Baker Tilly's Advanced Business Technology Services team to help them build and deploy a generative AI application using state-of-the-art natural language processing technology.

Strategy and solution

Success hinged on close collaboration. The teams joined forces to map business needs, pinpoint data sources and define key use cases. A crucial challenge arose: the knowledge base existed in a multitude of formats, from conversational transcripts to online help documents (HTML). Baker Tilly's Advanced Business Technology Services team brought their deep expertise to address this hurdle. The team crafted a custom solution called Retrieval Augmented Generation (RAG) powered by an AI agent. This intelligent agent, hosted on a secure and reliable platform like AWS Bedrock, could retrieve relevant information from these diverse data sources, regardless of format. The retrieved information then fueled the generation of natural and accurate responses to customer inquiries. Finally, the solution was seamlessly integrated with the Sage Intacct team's existing customer service platform, empowering them to monitor and continuously improve performance over time.

As a result:

• The project implemented a powerful and scalable generative AI application. This application boosted the efficiency and quality of customer service while fostering continuous product learning within the teams.
• The solution included intelligent search functions that locate the best answers to basic and complex customer questions. This empowered lower-tier support personnel to resolve these complex issues independently, reducing reliance on senior management.
• Finally, the self-learning capabilities of the model saved significant time and money on training and updating documentation. The AI can learn automatically from new data and feedback, eliminating the need for frequent training updates and cumbersome documentation revisions.