University of Oregon Partners with Dialect-Based AI Startup SoabBox Labs

The Irish tech company’s software will enable the College of Education to better identify struggling readers of different dialects.

Share this article!

A research and outreach unit at the University of Oregon’s College of Education has announced a partnership with Irish voice tech startup SoapBox Labs, to integrate the company’s dialectical artificial intelligence technology into the university’s assessment tool.

A press release accompanying the announcement said UO plans to integrate SoapBox Labs’ technology into CBMskills, a program which pairs with easyCBM, the platform the College of Education uses to measure the progress of entry-level readers. SoapBox Labs’ system can accurately assess children based on their own unique speech patterns, accents, and dialects, leading the company to win tech certification nonprofit Digital Promise’s Prioritizing Racial Equity in AI Design product certification in 2022.

The announcement follows two other recent SoapBox Labs collaborations with educational publishers Scholastic and Imagine Learning.  Gerald Tindal, director of behavioral research and teaching at the University of Oregon, said in the press release the incorporation of SoapBox Labs’ technology into CBMskills “[A]llows us to much more effectively and precisely unpack where students are struggling, so we can intervene more strategically and more quickly.”

“This partnership underscores our commitment to revolutionizing language and reading development through technology, and together with the University of Oregon, we’re continuing to leverage voice technology to help teachers and students,” Martyn Farrows, CEO at SoapBox Labs, said in the release.

UO’s College of Education launched, easyCBM, 15 years ago to assess acquisition of reading skills and inform educators’ interventions to support struggling students. Since its release, the program has been used to deliver eight million education assessments, according to UO. CBMskills is a voice-driven assessment platform which records and assesses timed reading passages to track student’s learning progress in order to identify and intervene with struggling readers.

Using the SoapBox language engine, the assessment tracks a student’s progress through timed reading passages, making it easier for educators to accurately and specifically identify where students are making mistakes in oral language proficiency. For example, a child who skips words or sentences but pronounces the words correctly may have a visual processing problem rather than a reading problem. If a child regularly replaces an “R” with an “lL,” an articulation disorder may also be present. The type and frequency of errors can be highlighted through a voice-enabled learning tool, enabling more personalized instruction and faster intervention.

The College of Education’s behavioral research and teaching division plans to onboard an additional 10,000 teachers with CBMskills by the spring of 2024, according to the announcement.