Using an AI Agent to Facilitate Student Data Literacy and Exploratory Data-Based Investigations for All

Handouts

Presented at 9:00 am in Mattie Silks on Friday, November 21, 2025.

Speaker(s)

• Geoff Freed, Director, onfocusAccess Consulting
• Daniel Damelin, Senior Scientist, The Concord Consortium

Session Details

• Length of Session: 1-hr
• Format: Interactive/Discussion
• Expertise Level: All Levels
• Type of session: General Conference

Summary

The Concord Consortium, in collaboration with Perkins School for the Blind, is developing an AI-powered agent called DAVAI (Data Analysis through Voice and Artificial Intelligence). This agent will be embedded as a plugin in the Common Online Data Analysis Platform (CODAP). DAVAI interprets verbal and typed commands from blind/low-vision users to perform data transformations in CODAP, generate data representations, and provide text and sonified descriptions of those representations.

Abstract

Data science has become essential in modern society, with growing career opportunities and widespread adoption in educational curricula. However, blind and low-vision students are significantly underserved in this field, often lacking the tools necessary for meaningful engagement with data.

With funding from the National Science Foundation, The Concord Consortium, in collaboration with Perkins School for the Blind, is developing and researching an AI-powered agent called DAVAI (Data Analysis through Voice and Artificial Intelligence). This agent will be embedded as a plugin in the Common Online Data Analysis Platform (CODAP), a free, web-based tool designed for exploring and analyzing data, particularly in educational settings.

DAVAI acts as a conduit between the learner, CODAP, and an AI agent, enabling it to perform actions for the user while also describing graphs and their underlying datasets, allowing users to conduct their own interpretation and data exploration.

Data descriptions are provided in both text and through sonification of graphs, allowing users to hear the shape of the data if they choose. This provides another dimension for learning about and understanding graphs. Using the CODAP API, the plugin retrieves the relevant graph data and maps the data to pitch and time values. The plugin then uses an audio framework to play those pitches in sequence within the browser.

Blind and low-vision high-school students are the primary audience for DAVAI, but we envision that the plugin will be useful for all learners, including students with physical disabilities that affect interaction through mouse and keyboard, as well as learners with cognitive disabilities who prefer using voice to control software and to hear descriptions of graphs.

Keypoints

1. AI agents have potential to allow blind and low-vision students to investigate and analyze data.
2. Sonification gives students an alternative method to explore data representations.
3. AI agents can benefit students with mobility and cognitive disabilities, as well as students who are blind.

Disability Areas

Cognitive/Learning, Mobility, Vision

Topic Areas

Alternate Format, Artificial Intelligence, Assistive Technology, Other, Uncategorized, Web/Media/App Access

Speaker Bio(s)

Geoff Freed

With 40 years of accessibility experience, Freed is a senior advisor to organizations working to ensure their digital experiences are usable by people with disabilities and are compliant with accessibility standards and regulations. Geoff helped lead digital accessibility efforts at the WGBH Educational Foundation and the National Center for Accessible Media (NCAM) for 35 years, before serving as the director of digital accessibility consulting at Perkins Access, a division of Perkins School for the Blind.

Daniel Damelin

Daniel Damelin has worked in the field of education for 30+ years, as a teacher, curriculum and technology developer, professional development leader, and educational researcher. Currently a Senior Scientist at the non-profit Concord Consortium, he has worked on numerous grant funded projects that resulted in open educational resources to support the learning of science and math through simulation and data visualization, inquiry-oriented curriculum and pedagogy, and formative assessment.

Handout(s)

• Using an AI Agent to Facilitate Student Data Literacy and Exploratory Data-Based Investigations for All