About Enlighten AI
What it looks like in action
What the students do:
Students write independently in the Enlighten platform and submit for multiple submissions rounds of AI feedback. They revise their writing based on feedback aligned to the teacher's rubric. Students can compare their score progression across submissions and compare scores with peers.
What the AI does:
Once the teacher has calibrated the feedback using a rubric, prompt, and reference materials, the AI provides immediate, rubric-aligned scores and categorized feedback to students on their writing. It helps standardize scoring and offers specific rubric-aligned suggestions. It also surfaces class- and school-level trends across submissions—common growth areas, patterns across classrooms—that teachers and leaders can act on.
What the teacher does:
Teachers upload a rubric, prompt, and reference materials and, over a few submissions rounds, calibrate the AI on their voice, grading style, and standards. From there, they can configure the AI to give students immediate feedback or hold it for teacher review first. As students work, teachers can monitor a live dashboard to spot which students need support and where common errors appear, then generate mini-lessons on widespread issues and structure discussions around the feedback.
What the leaders do:
School or system leaders can calibrate the AI to a shared curriculum or set of expectations if they've deployed it across classrooms for consistent instructional quality. From there, they can use school- and system-level trend data to identify where to focus coaching and professional learning.
Reach
500,000+
Students
(Cumulative, all-time.)
14,000+
Teachers
(Cumulative, all-time.)
2,000+
Schools
36
Districts
Use
Recommended use
Two assignments per week—one where students get AI feedback immediately to revise in the moment, and one where the teacher reviews and adjusts the AI's feedback before students see it; ~30 minutes per assignment.
Actual use
TBD — pending data.
Product dimensions
Subject
All subjects with rubric-based writing (ELA, Social Studies, Science)
Grade span
3–12, Higher education
Primary user
Teachers and students
Curriculum alignment
Standards only—teachers/leaders calibrate the AI on their own rubric, prompt, and curriculum content
What products are designed to do
Design scope
Specific instructional job
Teacher guidance
Nudges specific teacher actions (e.g., dashboard prompts during use)
The way AI interacts with students
- The AI asks questions and presents prompts students respond to
- Rubric-scored or written feedback (beyond right/wrong)
- AI gives feedback directly to students
Fit and adoption
Use mode
Solo by default — primarily independent, small groups possible
Subject focus
ELA/Literacy focus
Grade span
3–12, Higher education
Primary user
Teachers and students
How it works with curriculum
Standards only—teachers/leaders calibrate the AI on their own rubric, prompt, and curriculum content
Teacher override to AI scoring
Teacher can override the AI scoring
Product adoption
Both — teacher-adopted and system-adopted
Dashboard scope
Teacher dashboard 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
Enlighten gives feedback to all students during writing time enabling teachers to focus their attention on students who need more support.
“For me to get to 10 kids within 30 minutes and give them real feedback that's targeted to exactly what they would be doing, I would only probably be able to hit on some kids, but we felt like this program…would help support them getting feedback instead of just sitting and waiting for us.”
— Teacher, 4th Grade ELA
The live dashboard turns individual feedback into class-wide teaching points—teachers use trends to surface mini-lessons.
“Enlighten does a really great job of giving me a general idea—in the moment—of what everyone, the majority of the kids, need to work on. Based on what the dashboard is showing me, I'll show call something to show them because the dashboard also gives group trends, which is really great. It highlights teachable moments.”
— Teacher, 4th Grade ELA
Multiple submissions and visible score progression motivate revision—teachers describe students re-engaging with their writing because they can see growth in real time.
“Enlighten does a really great job of also being able to submit more than once so then getting feedback more than once and they can actually go back and see their old submissions.…They can see the progression after adding a little bit more each time...they're like, 'Oh my gosh, I grew five points. Oh my gosh, I grew, like, 10 points.' I think seeing that instant gratification for them means a lot.”
— Teacher, 4th Grade ELA
Limitations
Feedback vocabulary is often too advanced for below-grade-level and multilingual students—it works best for already-strong readers; teachers report translating or rephrasing the feedback for others.
“Based on the language or vocabulary in the feedback, some of the kids aren't able to read or understand the things that it's saying….The vocabulary words, if it were a bit more student-friendly or match[ed] individual students' reading levels, they would be able to internalize the feedback.”
— Teacher, 4th Grade ELA
Active teacher monitoring shapes how well it works.
“I'm always circulating….I'll pause them during intervals and then I will circulate and go to the kids who I know need more support and ask if they understand the feedback...I'll break it down. I'll also go back to the feedback and make adjustments because I know what they can do, and I know what they're producing, and I know what they mean, and sometimes Enlighten doesn't account for the nuances.”
— Teacher, 4th Grade ELA
Source-text keyword copying can register as strong responses—students copying source-text keywords can score high without producing real responses; the tool does not reliably catch this pattern.
“When kids just cop[ied] and pasted from the text, Enlighten was giving them a lot of points for it because they were using the words that were in the text… but technically it wasn't in their own words.”
— Teacher, 4th Grade ELA
Our instructional take
Instructional wins
Grading student writing is a heavy lift for teachers, limiting how much writing teachers assign and the amount of feedback students get. Enlighten addresses these bottlenecks so that teachers can increase the volume of writing students do as well as the volume of personalized feedback they get.
Immediate feedback is typically more actionable than delayed feedback. Enlighten shrinks the distance between the first attempt, feedback, and the second attempt.
The highly configurable rubric and calibration approach allows teachers to closely match the rubric expectations to the context of their instruction and the focus for writing they want to set up for a given assignment.
Instructional limits
The configurable rubric gives teachers flexibility but also creates room for inconsistent use and expectations across classrooms.
The AI-feedback on a rubric tends to favor a formulaic structure to writing.
Enlighten’s feedback can be overwhelming to students—leading students to give another attempt without processing the feedback.
All AI products, particularly writing focused products, are inconsistent in their ability to catch when students use other AI to generate their initial entry.
Implementation considerations
Enlighten 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, expectations for use.
Leaders need to think through whether to use a consistent rubric across classrooms or encourage teachers to use their own rubrics as well as whether to calibrate feedback centrally or have each teacher calibrate individually.
Leaders need to think through their expectations for teacher-student engagement during writing activities.
Evidence of Impact
Available evidence
Study underway
No published causal studies of student outcomes yet. EnlightenAI is a 2025 grantee of Accelerate's Call for Effective Technology program (DC, NJ); under that program, an independent research partner assigned by Accelerate conducts the evaluation. Findings are expected to be public following the 2025–26 implementation year. This is a quasi-experimental control, examining whether use of Enlighten supports student writing growth in AP U.S. History classrooms.
Sources: Accelerate Partners Hub (CET grantee listing) · Founder Convo (study specifics) · AI in Action Evidence Scan (May 2026)
From the Developer
From Enlighten AI: Where it's heading
We've added a differentiation feature so teachers can, by student, enable adjusted reading levels for feedback, text-to-speech, and translation. You can find an overview of our differentiation features here. We've also implemented the following additional updates:
Additional updates
Student grouping based on common skill gaps, so teachers can more quickly identify patterns and provide targeted support.
Differentiated reteaching resources aligned to the skills groups most need to practice.
Handwritten student work support, expanding the types of writing and assignments that EnlightenAI can support.
Student work portfolios that help teachers, students, and leaders track writing growth over time.
Usage analytics for administrators, including assignment volume, student revision activity, growth over time, and visibility into how teachers review and adjust AI-suggested scores.
Expanded LMS and rostering integrations to reduce setup burden and fit more smoothly into existing school workflows.
Product website: enlightenme.ai
Hear from the developer
Instruction Partners' CEO Emily Freitag spoke with Gautam Thapar from Enlighten AI to understand what problems they're trying to solve, how Enlighten AI works in classrooms, and what they're still figuring out.