OKO is a small-group math tool that uses AI to facilitate student discourse. Students try a problem on their own first, then talk it through as a group to agree on an answer, while the AI follows their discussion, poll responses, and answer submissions in real time. Teachers get real-time insights on how each student is doing on the math and how they’re collaborating with their classmates.
Students work in small groups of two to five. They can sit together around a table or a cluster of desks to collaborate in small groups. Students start each problem with a short solo attempt, then talk it through with their group—pointing out their thinking, comparing their written math work, and engaging in rich math discourse facilitated by the AI—to reach a shared answer.
The AI presents standards-aligned math problems, drawn from Illustrative Mathematics and Eureka content, and follows each group's discussion, poll responses, and answer submissions in real time. It prompts students to explain their thinking, prompts quieter students to join the conversation, asks students who are racing ahead to slow down and explain, and surfaces hints only when a group is genuinely stuck. During team time, it facilitates discussion rather than focusing on whether answers are right or wrong. It reads problems aloud and provides an on-demand glossary for language access, provides verbal prompts and cues, and scores each group on a four-level collaboration scale (i.e., Practice Squad, Free Agents, Dream Team, and Big Leagues) alongside a math proficiency score, giving the group teamwork feedback with specific improvement tips during the session. Afterward, it generates a session report for the teacher that details each student's math proficiency and collaboration performance, a problem-by-problem breakdown, exemplar quotations from the discussion, common misconceptions, and suggested next steps.
Teachers decide where OKO fits in the block (e.g., pre-teaching, reteaching, Tier 1 practice, extension). Teachers set up groups and push a problem set to them, choosing the standard and level for each session. Teachers can circulate to support individual students or pull a small group. They can adjust or override OKO's suggested groupings based on what they know about students' personalities and confidence; assign different groups different content to differentiate, pushing some students into above-grade-level material and pulling others back to fill foundational gaps; and review the session reports to plan reteaching, set class goals, and identify specific students to praise. Teachers can use real-time reports from the dashboard to intervene if a team gets stuck or conflicts arise.
School and system leaders decide to adopt OKO at the classroom, school, or school system level. They can set the scope of where it is used, from Tier 1 instruction to Tiers 2 and 3 intervention blocks.
11,820
Students
88
Teachers
30
Schools
17
School systems
Two to three times per week, 15–20 minutes per session (i.e., 30–45 minutes total per week)
Subject
Math
Grade span
3–12
Curriculum alignment
Curriculum-connected: Content is built and served in alignment to the scope and sequence of high-quality math curricula, starting with IM 360, Eureka Math, and Eureka Math² with additional curricula supported on demand
Design scope
Specific instructional job
The way AI interacts with students
Teacher guidance
Nudges specific teacher actions (e.g., dashboard prompts during use)
Social vs. solo use
Designed for small group use
Teacher override
Teacher cannot override—auto-scored
Product adoption
System partnership (districts or organizations buy and deploy across classrooms)
Dashboard scope
Teacher and leader dashboards available
Dosage recommendations
Provides dosage recommendations
There are clear trends in what users (i.e., teachers, students, instructional leaders) see as the wins and limitations of each product. For each trend below, we've included one illustrative quote from our user interviews.
OKO turns solo screen-work into social, structured math talk—students reason through problems together and build habits of discussion like listening, building on others' ideas, and working toward consensus. Teachers report better in-class discussion, and students practice communication skills that carry beyond math.
"After I started using OKO last year, I noticed significant increase in the kids being willing to talk in class and participate. It's not just them on a screen. They're talking, and it’s about the actual math; they’re learning how to come to consensus and collaborate….They're learning all these great communication skills that are, you know, transferable. It's so good for that."
— Teacher, 5th Grade Math
OKO facilitates the discussion within each small group (e.g., prompting, offering hints, keeping the talk going), so the teacher doesn't have to facilitate every group's conversation alone—freeing the teacher to move between groups and work with individuals.
"It helps me with small-group practice, and it offloads a group from my load that day and then I can pull one-on-one or I can pull a group up to the whiteboard, so it allows me more freedom."
— Teacher, 7th Grade Math
OKO supports multilingual learners' access to the math through read-aloud, a built-in glossary, and the chance to hear peers reason through problems.
"There's a good combination of visuals and voice—the AI will read it to them. There's a glossary in there that they can click on words that they don't know. And then they get that peer support as well. So it's really supportive for the English learners."
— Teacher, 6th Grade Math
OKO facilitates discussion but doesn't catch math errors in real time, so teachers must stay actively engaged—circulating and watching for misconceptions rather than treating it as a hands-off station.
"Team time is strictly just team time…it doesn't give in-the-moment feedback. And so sometimes students are kind of led down the wrong path.
— Teacher, 4th Grade Math
OKO structures the discussion, but the quality depends on the group's dynamic and on working at the right level of challenge. Confident students can dominate while quieter ones hold back, so teachers have to set clear expectations for collaboration, build students' habits of discussion, and keep each group on content that's challenging but within reach.
"There are a few students in particular [who] are confident to a fault….They always dominate the conversation. There's a need for building that classroom culture, the habits of discussion so all students are able to collaboratively participate."
— Teacher, 4th Grade Math
The all-verbal group format doesn't fit every student out of the box—some need a way to actively take part without speaking right away.
"I've got some kids on the spectrum that can use this but they need a little bit more time to learn how to be a part. And so if you just have it where they can listen and still do the math, but maybe not have the ability to speak right away."
— Teacher, 5th Grade Math
There is no independent, causal evidence of impact on student learning yet.
WestED is conducting a randomized controlled trial of OKO that is looking at student collaboration as measured by independent observers and surveys, teaching time savings, and student learning over an implementation period of 11 weeks. Full detail and sources will be added as soon as they’re available.
Sources: WestEd RPP Network (OKO Pilot Study) · AI in Action Evidence Scan (June 2026)
The following was written by the team at OKO and has been edited for clarity and tone by Instruction Partners.
OKO's near- to mid-term roadmap centers on four themes.
Product website: https://www.okolabs.ai/
Instruction Partners' CEO Emily Freitag spoke with Matt Miller from OKO Labs to understand what problems they're trying to solve, how the tool works, and what they're still figuring out.