AI Summary Tool
Led the design of an AI Summary feature that supports 200+ nonprofits worldwide in reviewing complex grant records and documents.
⏰ Timeline
4 Weeks
👥 Team
1 Engineer + 1 Designer
📌 Role
Lead Designer
Overview
Fluxx users were required to manually review large volumes of grant records and lengthy documents, creating a slow, cognitively heavy workflow that hindered productivity and delayed decision-making.
I led the design of an AI Summary feature, defining the interaction model, user experience strategy, and integration within existing record views. By embedding AI assistance directly into core workflows while maintaining strict privacy and security constraints, the solution reduced document review time by 25% and increased AI adoption by 15%.
Building for users new to AI
At Fluxx, we support grantmakers worldwide, but AI is still a new technology for much of our industry. The challenge was building an AI tool that is both useful and easy to adopt while ensuring organizations feel confident integrating it into their policies.
Users had varying opinions and no clear consensus
While some organizations hesitate, others move forward, integrating AI into their workflows. In the nonprofit world, it’s less about chasing trends and more about ensuring new technology aligns with their mission, while still boosting internal efficiency to increase their impact.
“Before AI can help us scale impact, we have to understand its risks and make sure it doesn’t create new barriers.”
“AI isn’t here to replace our mission, it’s here to help us live it out more effectively.”
Main objective: designing for AI adoption
Guide users through interaction, design the tool to teach users how to engage with the AI naturally as they use it.
Create a comfortable, approachable UI, develop a clean, simple interface to reduce intimidation and build confidence, encouraging adoption within daily workflows.
Ensure data transparency by creating tooltips and in-app notifications to provide clear context on data usage, addressing privacy concerns and building trust.
Provide Admin permissions and settings to allow customers to opt in to using the AI Feature
From prototype to production in four weeks
The initial proof-of-concept designs received such strong user feedback that the team fast-tracked development to launch the feature ahead of schedule. You can see the feature in action on Fluxx’s blog.
Delivery outcomes:
Achieved 15% increase in AI adoption
Reduced customer churn
Reduced manual document review time by 25%, as reported by customers
Moved from concept to in-app prototype in under four weeks, accelerating feature development by 90%
Feature outcomes:
Provided quick, AI-generated summaries to help users assess application alignment with fiscal goals
Reduced manual workload by streamlining a traditionally time-consuming process for reviewing grantee applications
Preserved human oversight in the review process, aligning with customers’ core ethos of keeping a human in the loop
Served as an entry point for AI-hesitant users, allowing them to feel in control while exploring emerging technologies
Constantly Iterating
After successfully launching the initial AI summarization MVP, I focused on evolving the feature based on user feedback and real usage patterns, with the goal of improving adoption and embedding AI into everyday work.
Built a scalable system capable of summarizing data from documents, applications, organizations, and future data streams
Designed the interface to feel native to existing workflows rather than a standalone AI tool
Established a forward-looking foundation to continuously improve summarization as new data sources and AI capabilities are introduced
Integrate AI Summary into future, larger-scale features
Design work is continuous and ongoing, and continues to inform Fluxx’s broader AI roadmap.
