Artificial Intelligence in Alma
Artificial intelligence is accelerating, and transforming higher education, research and libraries’ work: with news of new technologies and capabilities on an almost daily basis, with more and more data being produced using AI, with new copyright questions constantly being introduced and with evolving patron expectations. Libraries are also facing immense challenges, including limited resources, expanding scope of materials, growing complexity of standards and patrons that expect the world in the palm of their hand.
Clarivate aims to help libraries navigate this new world, working with our community to build innovative solutions and tools that meet the forming library standards for AI. Alma’s AI solutions are based on Clarivate’s Academic AI platform, which aims to promote a responsible, effective and trustworthy use of AI. We recognize that AI is a complex statistics and probabilities engine, which cannot replace a human being’s understanding, common sense and morals, or a librarian’s expert knowledge – our aim is to build tools that help librarians and libraries, while maintaining control in the library’s hands.
Our Vision
We aim to introduce AI-in-the-loop solutions, making librarians’ work easier and more efficient, and allowing libraries to meet their patrons’ expectations.
Our vision for integrating AI in library management consists of two main themes:
Innovative Efficiency Tools
We aim to create AI tools that cut through challenging workflows and automate mundane processes, letting librarians focus their effort on the complex challenges that require their expertise. Our goal is maximizing automation and batch processes, and minimizing manual labor for librarians when appropriate.
Intuitive interface
We explore solutions for introducing more intuitive interfaces, allowing librarians to more easily interact with Alma, find what they need and perform actions.
Using innovative AI agent solutions, we aim to introduce a conversational experience, letting librarians who prefer it converse with Alma to get information and perform actions. This will be done by combining field-specific capabilities, according to that field’s data, workflows and needs, which will be combined into a single, familiar and user-friendly interface librarians will be able to work with according to their roles and access.
Our Principles
Ethics and Fair Use
We take great care to make sure our AI solutions are beneficial and responsible, and conform not only to legal requirements, but also to standards being introduced in the library world, such as the forming PCC Guiding Principles for Use of AI and Machine Learning Technologies in Cataloging and Metadata Work by the PCC Task Group on AI and Machine Learning for Cataloging and Metadata.
Responsible AI Principles, Informed by the Library Community
Our AI-powered solutions are developed in close partnership with our customers and community, through close conversations, collaborations, and our work with interested librarians in the Alma AI Focus Group.
Beneficial Workflows
As AI tools may have adverse impacts, we carefully consider the usefulness of introducing AI into Alma flows, using AI in ways that can meaningfully benefit the library and librarians.
Leaving Control in the Library’s Hands
The choice of whether or not to use AI in Alma workflows is the library’s, and our AI workflows maintain human autonomy, giving librarians the choice whether to use the AI flow.
Conforming with Industry Best Practices and Guidelines
Alma’s AI developments are informed by Clarivate’s Academic AI platform’s expertise and compliance with industry best practices and guidelines, to introduce AI tools in a secure and stable way.
AI in Library Workflows
Metadata and Cataloging
See the AI Metadata Enrichment for Libraries theme page for details on our plans for AI and metadata.
Alma Assistant
With constant growth in the scope of materials and resources handled by libraries, as well as the complexity of selection and collection management, we are working with our community to identify areas with potential for innovative workflow streamlining and improvement. We are undertaking initiatives to harness advanced technologies to assist Alma staff users in their day to day work, with intuitive assistants that will allow staff to ask natural language questions and easily perform actions.
First area: Acquisitions' Funds and Ledgers
The first focus area for the Alma Assistant will be Acquisitions – Funds and Ledgers. Staff users with the appropriate roles will be able to ask the Assistant questions about their institution’s funds and ledgers and receive immediate, natural-language responses. They will also be able to perform actions, such as updating fund information.
For more information, see Alma’s roadmap plans.
Alma Insights
This new AI-powered feature is designed to transform your analytics data into clear, accessible, and meaningful insights—delivered directly within Alma. It periodically highlights relevant trends and information across different areas of the library, helping all team members stay informed and engaged without the need to run complex Analytics reports or sift through raw data.
How it works?
- An OTB Analytics report is defined – containing meaningful data
- The product team defines:
- How often to generate insights on this report
- Which user roles should receive them
- Specific instructions to the LLM (prompt)
- When it’s time to generate an insight, Alma extracts the report data, and passes it to the LLM to analyze and provide insights.
- Alma displays the insights to all users with the relevant role(s) as a "notification". Users can download the report data the insight was based on and check it out:
For more information, see Alma AI Insights.
Analytics Assistant
Library analytics are critical for libraries, from analyzing and improving library data and workflows, to periodical reports to management showing the library’s effectiveness, understanding and representing the library’s statistics is key.
A new capability enables users to interact with Alma Analytics reports through a conversational AI interface. Instead of navigating complex report structures or filtering data manually, users can ask questions in natural language—such as “How many items were loaned last month?” or “Show me overdue fines by user group”—and receive instant, relevant results from existing reports.
For more information, see Analytics roadmap plans.