Case Study: a government organisation reduces customer query workload with Boost’s AI chatbot POC

a government organisation with a customer-facing role validates an AI chatbot in a 2-week POC with Boost

A government organisation with a customer-facing role was challenged by a high volume of repetitive customer queries, which reduced staff efficiency and left less time for specialized work. They wanted to explore an AI chatbot solution but were unsure if it could handle their complex information and provide trustworthy answers. They engaged vendor Boost to conduct a low-risk proof of concept (POC) to trial the idea.

Boost implemented a retrieval-augmented generation (RAG) chatbot for the POC, designed to provide accurate answers by only using the organisation's website content and citing its sources. The solution successfully answered common queries, demonstrating its potential to free up staff. The POC built trust within the organisation and provided valuable feedback for future improvements, validating the chatbot's usefulness and securing stakeholder buy-in for Boost's proposed future development.