Introduction

Annual reports are usually… well, a little boring. Think pages of text no one finishes, and data buried somewhere between page 3 and 103. But what if reading an annual report felt more like talking to someone who actually knows what you’re looking for?

That’s exactly what Enfection and People’s Leasing & Finance PLC set out to do. Together, they launched an interactive AI avatar agent for an annual report, ushering in a new era of intelligent and intuitive stakeholder engagement. Viewers are able to read the report and interact with it. 

The Project

The idea was simple but bold: Replace static, one-way communication with a dynamic, conversational experience.

Powered by Retrieval Augmented Generation (RAG) and built on large language models, this AI avatar was trained solely on PLC’s published annual report. From balance sheets to sustainability goals, it can instantly answer stakeholder questions with precision.

This solution was about solving a real problem: making dense corporate content more accessible, especially as the stakeholder landscape shifts. With more millennials and Gen Z professionals engaging with corporate reports, whether as investors, regulators or customers the need for smarter, more engaging formats has never been clearer.

Results & Impact

The AI avatar not only delivers data, it also builds trust, creates transparency and invites curiosity. Stakeholders can now interact with annual reports the same way they use AI in their everyday lives: naturally and on demand.

By leveraging AI’s speed and efficiency, the concept went from idea to live deployment in a short amount of time. Faster delivery, significant savings, and a stakeholder experience that redefined how annual reports can perform in the digital age.

This is more than a one-off innovation. It’s a model for how companies can use AI to elevate not just what they say, but how they say it. The success of this project opens doors to broader applications, from HR communications and compliance to customer support and internal knowledge systems.