The goal sounded simple: make email communication smarter and more efficient for customer service teams. On closer look, however, you start to realize how many variables there are. We needed to create an AI experience that agents are comfortable using, with accurate content straight from their knowledge-base, adapted to different languages, cultures and scenarios.

Building this AI tool required us to carefully consider all of these factors. It required an understanding of the back-end hardware as well as the limitations we had on the front-end, utilizing our legacy email platform. In order to familiarize myself with the back-end content, I studied AI, receiving two credentials from Oracle Cloud Infrastructure, and partnered with key clients that were open to experimenting with and helping out with our pilot. Together we created a v1 that generated emails based on topic, making it easier on the implementation side itself. This experimental v2 is more agent-prompt driven and allows them to add in information that may not be immediately obvious to a copilot tool.

Based on our takeaways from the pilot itself and from un-moderated user testing, we were able to create a multi-faceted email tool that allowed for regeneration, simplify and custom prompting. We are currently doing further research into what expand could be and we're learning that configuration of a lot of these features depends on how strong your prompting is. Later adaptations will include a wizard to get started with AI and guidance around prompting to help make this easier for administrators to stand up.