How a VEO Uses Your Organization’s Voice to Engage Donors

One of the most common questions our team hear from fundraising teams is: How does the VEO know what to say?

When a Virtual Engagement Officer (VEO) communicates with your donors - sharing impact updates, sending invitations, or making solicitations - it can’t just sound like a member of your team. It needs to communicate as a member of your team, reflecting the right tone, priorities, and strategy.

Authentic engagement starts with the right inputs - fundraising intelligence and your institution’s unique voice - working together to guide every message. It takes great fundraising expertise and engineering skills to accomplish this.

At Givzey | Version2.ai, we start with an Organizational Knowledge Base, built from your organization’s own content (e.g. website, fundraising and marketing emails, newsletters, event calendars, social posts, and the various channels you use to communicate with your community). These materials ensure the VEO knows exactly what’s happening throughout your organization, can prioritize the most recent and relevant content, and do exactly what a human gift officer would do: decide what is most relevant to each of the donors in the portfolio and use that content for meaningful individual and personalized cultivation, solicitation, and stewardship.

What makes this powerful is how the content is structured and used. During onboarding, we ingest materials from multiple formats and convert them into markdown, a format designed to be LLM-readable, with added metadata -  including content type, source, event dates, and tags. This makes it easier for the system to interpret meaning and relevance.

All content is stored securely in Amazon S3, tied to your organization. From there, it’s indexed via Amazon Bedrock into a vector database. Unlike a CRM or SQL database that searches by fields and tables, a vector database allows the VEO to search based on semantic similarity - meaning it retrieves content based on what it means, not just what it says.

When the VEO writes to a donor, it builds a custom prompt based on donor profile, message type, channel, and strategy. It then queries the vector database to retrieve the most contextually relevant content - like a campaign update from your newsletter, a Giving Day reminder from your email archive, an event pulled from your calendar, or otherwise.

Why is all of this so important? Donor trust is at the forefront of everything we do. Nothing is more important to my team and me than getting it right with your donor. The organization-specific secure knowledge base provides the necessary guardrails for the VEO to only use your existing and approved content. You control what goes in, and by extension, what the VEO can see and say.

The technology behind this level of trust is truly amazing. We invite you to hear directly from Givzey | Version2.ai’s Bill Dusch, PhD. Bill is a machine learning engineer who joined a recent webinar to walk you through the intricacies of solving these challenges so organizations can use digital labor to add trusted fundraising capacity to their teams.