I am currently leading a project with a dedicated three-engineer team, focused on refining FreeToBook's client ticket support system. Our primary goal is to reduce response delays for routine inquiries without compromising the high-quality, personalised nature of their expert support.
Rather than building a client-facing chatbot, we engineered an internal, LLM-powered assistant designed to empower the existing support staff. The system analyses incoming tickets, flags critical context, and drafts a proposed initial response. By continuously learning from past responses, edits, and feedback, the system compounds its effectiveness over time. This approach allows the FreeToBook team to concentrate on more complex and unique challenges while drastically speeding up the handling of common, repetitive questions.
The results have been significant: in the first six months of operation, this internal tool drove a 22% reduction in the median reply lead time. I am working on this project in a fractional capacity, dedicating a couple of hours per week. My role focuses on setting the strategic direction, upskilling the internal developers, building foundational prototypes, and guiding the team through data-driven architectural execution.
