Kicking off the year with our customers: Customer Advisory Board meeting dives into GenAI

A semi-annual tradition, IntelePeer’s Customer Advisory Board (CAB) gathered in January to discuss relevant topics concerning today’s business technology landscape. These executives bring unique perspectives from an array of industries, such as healthcare, mortgage services, utilities, FinServ, FinTech, and insurance, as well as a major North American gymnasium chain.

In the first meeting of the new year, the CAB examined artificial intelligence (AI) and Generative AI (GenAI), its impact on business, and included best implementation practices to ensure the highest quality customer experience and data safety. Here are some highlights.

Roundtable discussion: General AI

CAB members investigated several AI topics – namely, co-pilots, data and AI, channels and modality, and best AI use cases.

• AI co-pilots: A conversational interface that employs large language models (LLMs) to aid knowledge workers and customers in making decisions and completing tasks. These context-aware assistants can respond dynamically to inquiries and requests.
• Data and AI: An AI-powered solution is only as efficient as the data it has at its disposal. Organizations eager to deploy GenAI must first outline their data needs and decide which consumption approaches they will use.  
• Channels and modality: Brands should determine which type of customer communications channels would benefit from a GenAI integration. Ideally, companies should adopt a multi-modal approach that uses voice, text/chat (rich content), forms, and video.
• Best AI use case: Companies must identify those use cases or business areas that would yield the greatest productivity and cost-efficiency if automated through GenAI.  

Intent Studies

Beyond those general AI topics, CAB members also delved into other advanced AI applications, like IntelePeer’s Customer Intent Study. IntelePeer analyzes up to 20k calls per customer over a period of two weeks in a secure manner, stripping any Personal Identifiable Information (PII) and Protected Health Information (PHI) at the source and anonymizing customer data.

The studies help businesses learn why people reach out in the first place and the data is then used to define and prioritize those main interactions that can and should be automated. Rather than guessing where to put AI in the customer service environment, businesses now have the data they need to construct an implementation roadmap. In essence, the data collected from an Intent Study fuels customer automation.  

To date, IntelePeer has helped customers in many verticals, including healthcare, dental, retail, insurance, fitness and logistics. CAB members who have been through the process expressed support of the approach and commented on the valuable insights they gained from the study.

Response guardrails for generative AI

Another key moment from the discussion was focused on the safeguarding of data. A haphazard or improper GenAI deployment can result in output quality issues, hallucinations, copyright infringement, and biased algorithms. At the same time, mismanaging data could lead to accidental exposure of propriety or sensitive information, having considerable legal implications (Forbes, 2023).

As such, the CAB discussed how businesses need to have response guardrails when applying GenAI in the customer service environment. IntelePeer utilizes retrieval augmented generation or RAG, which feeds data context to an LLM to get responses grounded in a customer-provided dataset. Throughout the entire process, from the moment the data gets prepared until the LLM sends a response to the client, the necessary guardrails prevent any sensitive information from being exposed.

Q&A patterns and RAG

Question and answer (Q&A) patterns also emerged as a noteworthy topic during the CAB meeting. IntelePeer uses Retrieval Augmented Generation (RAG) for Q&A patterns, which begins when a customer asks a question to an AI-powered bot. The bot performs a lookup of the question in the database – if it can’t find the question and corresponding answer, it will consult the knowledge base. If it still cannot find an answer, it will transfer to an agent and save the question to the Q&A database.

Later, a human will review this new question, conduct a dataset import, and save the answer to the knowledge base. Ultimately, no exceptional question goes unanswered. With the RAG process in place, businesses can maintain control over response sets for interaction automation.

Thank you to those who participated

IntelePeer would like to thank our CAB attendees for generously giving their time, energy, and knowledge to another productive and elucidating discussion. With each successful meeting, IntelePeer gains invaluable insights into the challenges and issues affecting businesses today, allowing us to create truly innovative solutions.

Contact us today to discover more about IntelePeer’s AI-powered Customer Interaction Intent Study.