AI made practical
Making AI Work for Library Staff.
The challenge
Making AI accessible, practical, and effortless for library teams.
AI tools like ChatGPT and MidJourney offer huge potential for libraries. From drafting communication messages to generating visuals and refining program ideas. But in practice, many library staff members aren’t yet using these tools to their full potential. Reasons include a lack of knowledge about the tools’ capabilities, uncertainty about how to apply them to specific tasks, and limited consideration of AI adoption from an organizational perspective.
How we solve it
AI support designed around real library work.
We put UX at the heart of this project. By interviewing library staff, we identified where AI could truly support daily tasks. These insights shaped user stories and led to Bibliobot — an intuitive interface built over AI tools like ChatGPT and MidJourney.
The result is a user-friendly assistant designed for real workflows, no technical skills needed. The first version helps staff quickly set up activities and programs, making daily work smarter and simpler.
Quick sketches to share initial ideas.
With the kick-off brief and a strong collaboration with the development team we established initial thoughts on:
- Desired user flow.
- Possible components.
- Prompt and user inputs.
What are we trying to build?
We create a roadmap—a strategic plan for product development that outlines which features will be implemented now, next, and later. It serves as a tool to align the team around:
- Goals
- Priorities
- Product mission
From wireframes to prototype
We dive into the design process, collaborating closely with developers and stakeholders, iterating on:
- Test with users and iterate based on feedback
- Wireframes to visualize the user flow
- Iteration with developers and stakeholders
- Prototype for user testing
Developer aligned documentation
Developer-Aligned Documentation
We align with the team on documentation preferences to ensure clarity and consistency, including:
- Sizing across screens
- Behaviours
- Naming conventions
- Availability to follow up and beta testing