Atlas by Kiddom
Atlas by Kiddom is a math tool that uses AI to turn each day's end-of-class assessment into targeted practice, intervention, or enrichment during the next lesson. As students finish, it grades their work, shows the teacher the misconceptions behind incorrect answers, and builds every student a differentiated warm-up from the curriculum to start the next day’s lesson.
What it looks like in action
What the students do:
Students complete their math work as usual and complete a short end-of-class assessment on the platform, consisting of a few multiple-choice questions tied to the day's lesson. They can work the problems out on paper or using digital tools, then enter their responses. Following class, students receive teacher-provided feedback on the day's activity. As a warm-up, students complete a short activity tailored to where they are: students who missed problems work through related exercises targeted to the misconception; those who got everything correct move on to extension activities. Students then complete the day's lesson, guided by the teacher.
What the AI does:
As students submit their answers, the AI grades them and identifies, in real time, the misconceptions behind incorrect answers. The AI connects each misconception to where the underlying skill is taught in the Illustrative Mathematics curriculum, including pulling up material from earlier-grade lessons, and generates ready-to-use activities for the next class: a warm-up or do-now, small-group reteach, or extension work for students who already mastered the material. It provides guidance to teachers to group students by the misconceptions they share, and for each group, it writes a short description of what that group misunderstood, a guiding question, and the activity to work on (that the teacher would lead, there is no technology in this portion of the lesson). The AI also drafts feedback on students’ incorrect answers for teachers to approve, edit as needed, and deliver. Feedback is designed to prompt a student's thinking rather than give the answer. The AI runs within a closed system limited to Illustrative Mathematics materials, so, unlike a general-purpose AI chatbot, it can't pull in external content or answer off-topic questions.
What the teacher does:
The teacher runs the day's Illustrative Mathematics lesson and assigns the end-of-class assessment called the cool-down. While students work, the teacher can see a live feed of how the class is doing and which misconceptions are showing up. They can choose to step in to support specific students in the moment rather than waiting to reteach the whole class the next day. Teachers can ask Atlas what to do next and get suggested instructional steps, not just data to interpret on their own. Before the next class, the teacher reviews the misconception summary, the suggested student groups, and the activities built to address each group's needs. The teacher can decide to: keep the groupings based on what they know of their students; choose whether to run the reteach as a small group, a do-now, or a whole-class review; and edit or reject any activity. They also review the feedback the Atlas drafts for incorrect answers and can send it, edit it, or hold it. The teacher provides the feedback to the students, not the AI.
What the leaders do:
Once leaders adopt Atlas as part of their core curriculum, they can decide how it fits within their instructional program. Across schools or classrooms, leaders can see which standards and misconceptions classes are struggling with and how consistently teachers use Atlas through school system dashboards. They can pull the reports they need from their own data, track how students are likely to perform on assessments, and then direct coaching and support to the teachers and schools that need it. Leaders can also query the data using natural language to understand patterns, glean insights, and determine where support is most needed.
Reach
~8,000
Students
-
Teachers
29
Schools
1
School systems
(District 11, NYC); broad availability beginning Fall 2026
Use
Recommended use
Three times per week
Actual use
TBD—pending data
Note: Students complete a short multiple-choice cool-down at the end of each Illustrative Mathematics lesson, which feeds the next day's AI-generated intervention materials.
Product dimensions
Subject
Math
Grade span
6–8
Curriculum alignment
Curriculum-embedded: HQIM-integrated with Illustrative Math®
Design scope
Specific instructional job
The way AI interacts with students
- AI does not interact directly with students
- Right/wrong signal (most of the time)
- Feedback routes through the teacher
Teacher guidance
Nudges specific teacher actions (e.g., dashboard prompts during use)
Social vs. solo use
Mixed (platform)—some uses social, some solo
Teacher override
Teacher can override the AI scoring
Product adoption
Both—teacher-adopted and system-adopted
Dashboard scope
Teacher and leader dashboards available
Dosage recommendations
Provides dosage recommendations
Users' take on wins and limitations
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.
Wins
Atlas analyzes student responses in real time, so teachers can respond to misconceptions while there’s still time in class.
"I literally get feedback instantaneously…'eight students correctly calculated the IQR [interquartile range] but confused the meaning of IQR with range'….I can say, ‘hey, let's talk about IQR again and fix the issue before it becomes a lasting issue.’"
— Teacher, 6th Grade Math
Atlas does the sorting of student work a teacher would otherwise do by hand after class: dividing the class into groups and pairing each with the right activity. Instead of spending prep time building those groups from a stack of exit tickets, the teacher walks into ready-made groupings and can spend that time teaching them.
"Normally when I give paper cooldowns, I look through them and separate them into different piles of like, ‘ok, you understand, you made a misconception, you're almost there, and then you have no idea what's happening’….That takes so much time after school, whereas Atlas, as soon as the students answer, everyone has [their group]."
— Teacher, 7th Grade Math
Rather than only reteaching today's lesson, Atlas can locate the earlier skill a misconception actually stems from, sometimes several grades back, and pull that specific content forward, so the reteach targets the root of the gap instead of its symptom.
"We were working on circles, and one of my students' misconception was a standard from 4th grade. And it incorporates that to say, ‘ok, the reason the student got this incorrect is because it may have misconceptions around this 4th-grade standard.’"
— Teacher, 7th Grade Math
Limitations
There is so much content available to teachers that it can be hard to get through it all.
"For students not yet at grade level, I'll kind of have to do more of the hand-holding, or go just a little bit slower, just to make sure everyone is engaged and I’m meeting everyone where they're at. I have to use a limited version of Atlas…mostly stick to the warm-ups.""
— Teacher, 7th Grade Math
The feedback Atlas drafts often gives away more than a teacher wants a student to have—enough to solve the problem for them rather than prompt them to rethink it.
"Sometimes it's too much feedback, where…I still want them to struggle a little bit….I would change the AI feedback to something like, 'go back and rethink about this question'….I may only take a piece of what the AI was giving me."
— Teacher, 6th Grade Math
Atlas’ feedback uses math vocabulary and phrasing that isn't accessible to every student, particularly for multilingual learners. So before it's useful, some teachers say they have to rewrite it in plainer language. The tool saves time on grading but hands some of it back as translation work.
"I understand the feedback and the feedback is great, but for some students, especially my ELs (English learners), they may not understand exactly what it's trying to say. So I have to go back and change some of the wording, just so they can understand exactly what's being said."
— Teacher, 7th Grade Math
Our instructional take
Instructional wins
- Atlas offers a light, high-leverage use of AI. Because it’s deeply embedded in the curriculum, it helps teachers analyze and determine the implications for the next lesson from the cool-down—a critical, complex instructional job that Atlas does more quickly and with more consistent efficacy than teachers can in paper-pencil instruction.
- Targeting misconceptions, in particular, is complex and important for math learning. The design of Kiddom’s cool-downs enables immediate identification of these misconceptions in a highly actionable way, making it more likely that students will receive instruction the next day that targets their core learning needs.
- The immediate feedback on cool-downs Atlas offers reinforces what students learned and clarifies where they still have work to do more precisely than they could get from paper-pencil exitexit tickets. This benefits student learning and saves time for teachers.
Instructional limitations
- Atlas identifies what to reteach and provides ready-made differentiated material that teachers can assign in one click, along with tools to give feedback and scores on the resulting work. Whether teachers assign that material, how well they deliver the reteaching, and whether students engage with it still depend on teacher judgment, time, and follow-through, so the impact on student learning will likely vary widely.
- Though the multiple-choice cool-downs are designed to flag misconceptions, the structure of the questions does not require students to show their thinking, making it possible for students to guess and for the technology to flag the wrong misconception, limiting teachers' visibility into what drove student responses.
Implementation considerations
- Atlas shares the standard school-based implementation considerations: scheduling (within the class schedule and the scope and sequence for instruction), teacher and leader training, tech set up, and expectations for use.
- Leaders need to think through how closely they want teachers to follow Atlas’s guidance and when they want teachers to have the freedom to deviate from the recommended structure or student groupings.
- Leaders should decide when and how Atlas data should be used in grading and during PLC and planning time.
Evidence of Impact
Available evidence
No evidence yet
There is no independent, causal evidence of impact on student learning yet.
Kiddom, in partnership with the Johns Hopkins Institute for Education Policy and Teaching Lab, is conducting a mixed-methods evaluation of Atlas designed to meet ESSA Tier 2. The study compares classrooms using Kiddom with Atlas, Kiddom with Atlas plus instructional coaching, and Kiddom without Atlas to isolate Atlas's contribution to student mathematics achievement. It also examines teacher practice, instructional decision-making, and implementation fidelity, drawing on state assessment results, interim assessments, curriculum-based measures, classroom observations, teacher surveys, platform usage data, and interviews. Results are expected in January 2028.
Source: Study details provided by Kiddom
From the developer
From
Atlas by Kiddom
The following was written by the team at Kiddom and has been edited for clarity and tone by Instruction Partners.
Atlas is evolving from a single intervention workflow into a more connected system for using student evidence during core math instruction. Recent updates include real-time synthesis of cool-down responses and a unified Atlas Activities workspace that keeps differentiated activities alongside the Illustrative Mathematics lesson. Teachers can choose among small-group and whole-class activities; individual practice and spiral-review are coming soon. Spanish support is now available for Atlas cool-downs. Integration with Kiddom’s paper scoring workflow and digital pacing guide are coming soon.
Product website: https://www.kiddom.co/atlas
Hear from the developer
Instruction Partners' CEO Emily Freitag spoke with Abbas Manjee from Kiddom to understand what problems they're trying to solve, how the tool works, and what they're still figuring out.