Project Vision
Bonterra Que is an AI-powered fundraising assistant designed to help nonprofit practitioners act faster and smarter without replacing their expertise. It brings together data, context, and actionable insights across Bonterra products, helping fundraisers prioritize tasks, optimize campaigns, and spend more time engaging supporters instead of wrestling with tools. Que proactively surfaces recommendations—from segmenting donors and coaching users to automating emails and forms—so fundraisers can make confident, informed decisions quickly.
My work focused on building a go-to-market version 1 of a platform agnostic chatbot and evolving the chat experience by introducing “object creation.” The goal was to let the LLM generate files that users would normally create manually. Users interact naturally through the chat, while the LLM handles the heavy lifting; managing the entire workflow within the chat interface. We'd be using donor segmentation as the base exploration and first go-to-market feature for Que.
Challenges
1. Work with a legacy team while pushing sustainable design changes
2. Consider adaptable design for a responsive web environment
3. Iterate effectively while adapting to rapidly changing and non-cemented standards
My Role
Senior Product Designer
Tools Used
Figma
ChatGPT
v0 by Vercel
Sharepoint
Zoom
Confluence
The Goal
Competitive Analysis
When designing Que, we looked at how other AI chatbots like ChatGPT, Claude, and Gemini to handle recommendations and guidance. This was helpful in understanding what other experiences our users may already be familiar with, as well as noticing what conversational features to look out for. Seeing how they provide context-aware suggestions and keep conversations intuitive helped us make design decisions faster and with more confidence. It also reinforced our approach: Que should feel like a helpful assistant that supports fundraisers, not replaces them.
Object Creation
As Que evolved, we wanted it to feel less like a chatbot and more like a true partner who could coach, guide, and take action on your behalf. From a simple natural language request, users can now create and modify “Cards,” flexible UI elements that represent real documents or assets within the platform. These Cards act as both output and entry points, allowing users to preview, download, or move into more complex workflows without leaving the chat. By blending conversation with creation, Que turns chat into a powerful workspace where fundraisers can think, act, and build in one place.
Preview Mode
For more complex tasks like building forms or email campaigns, Que offers a preview mode that lets users review, edit, and refine their work before finalizing it. This gives fundraisers space to collaborate with the assistant, making adjustments in context and ensuring the final output feels intentional and ready to use from within the chat experience.
Next Steps
Que launched in the Summer of 2025 to a roaring success. over 12% of our user base immediately started using the tool, with adoption rising month-by-month. Our support team has continued to gather feedback on what user needs are, informing our approach as we begin building out new skills for the AI.
A few areas of exploration:
- Donor Segmentation
- Donor Stewardship
- Drip Email Campaigns
- Individual Donor Profiles
& much much more!
What I learned
The challenges listed in this case study were no joke. It was difficult to balance working with a legacy team, influences from leadership, all while building out AI standards and adapting our design system to accommodate AI features. It was the closest I have worked with my design team, having nearly daily stand-ups for this single initiative along with hundreds of slack threads to manage and keep track of.
Through it all I learned how to adapt my personal process to optimize methodology to boost team moral and be an effective contributor, stepping in where I could to take the load off of team mates and fill process gaps.
