AI-powered teacher evaluations
for Marshall schools.
Cut evaluation documentation from hours to minutes. EvalScribe is built around the Kim Marshall Teacher Evaluation Rubric natively — its six domains and the Does Not Meet Standards to Highly Effective scale, in the rubric's own language, and built for the mini-observation workflow the rubric was designed around.
Saved per evaluation
The Marshall rubric, built in
Does Not Meet to Highly Effective
Evaluator-controlled ratings
Three evaluations free — no credit card · Individual plans from $100/year · School & district licensing available
Endorsed by educational leaders
Game changer. Absolute game changer for school admin.
Tobin Davidson · Assistant Superintendent · 40+ years in education
This is amazing. I would pay lots of money for this. I’m serious — it’s crazy good.
Kittie Brown · Assistant Principal
Also endorsed by Dr. Robert Stinson · Glenn Brown · Angie Brandt — read full quotes ↓
Built by a classroom teacher
Founded by a practicing K–12 educator with nearly 20 years in the field.
Marshall-specific
All six Marshall domains and the four-level scale, built for mini-observations.
Privacy-first
Transient processing on Microsoft Azure. Never used for AI training.
Built for workflow
Designed around the realities of frequent mini-observations.
The Kim Marshall rubric. EvalScribe is built around all six domains.
Developed by Kim Marshall, the Teacher Evaluation Rubric organizes teaching into six domains and fifty-four criteria, scored on a four-level scale from Does Not Meet Standards to Highly Effective. It is designed to be filled in from frequent, short mini-observations rather than a single formal visit. EvalScribe is built around the rubric's own structure and language.
Kim Marshall Teacher Evaluation Rubric
The Marshall rubric is a year-end assessment of where a teacher stands across every part of the job, meant to be informed by unannounced mini-observations every few weeks and the face-to-face conversations that follow. Its six domains cover the full arc of practice, from planning through professional responsibilities. EvalScribe maps your observation notes to the criterion they support and drafts a rating and evidence-anchored feedback in the rubric's own language.
Six domains. Fifty-four criteria.
Planning & Preparation for Learning
9 criteriaContent expertise, standards-aligned goals and units, assessments, anticipating misconceptions, and an organized environment.
Classroom Management
9 criteriaExpectations, relationships, respect, routines, prevention of problems, and productive use of class time.
Delivery of Instruction
9 criteriaClear objectives, engaging and rigorous lessons, connections, effective questioning, and a repertoire of strategies.
Monitoring, Assessment & Follow-Up
9 criteriaChecking for understanding, timely feedback, recognition, and using data to adjust and support every student.
Family & Community Outreach
9 criteriaRespect for families, clear and regular communication, involving families in learning, and responsiveness to concerns.
Professional Responsibilities
9 criteriaReliability, professional judgment, teamwork, openness to feedback, and ongoing growth and leadership.
The four Marshall ratings.
Does Not Meet Standards
Improvement Necessary
Effective
The expected standardHighly Effective
Marshall is deliberate about the scale: Effective is solid, expected professional performance, and teachers should feel good about it. Highly Effective is reserved for truly outstanding practice against very demanding criteria. EvalScribe drafts to each level's own criteria, so Effective isn't treated as a middling score.
Built for mini-observations
Marshall is explicit that the rubric is not a checklist for a single visit. It is meant to be filled in at year's end from many short, unannounced mini-observations, each followed by a quick face-to-face conversation. That cadence — observe often, give feedback fast — is exactly the workflow EvalScribe is designed to speed up.
The rubric is also open source and adaptable, so districts adjust it to fit. EvalScribe drafts the observation-based write-up; the overall summative rating is set through your district's process.
How EvalScribe handles the Marshall rubric
All six domains and their criteria, along with the four-level scale, are built into the tool, with the distinctions that separate Does Not Meet Standards from Improvement Necessary from Effective from Highly Effective. EvalScribe maps the evidence from your mini-observation notes to the criterion it supports and drafts best-fit ratings and evidence-anchored feedback in the rubric's own language — so the write-up is defensible, traceable back to what you observed, and ready to inform the year-end evaluation.
Scope of the score. EvalScribe drafts the observation-based write-up. The overall summative rating and any student-growth measures are determined by your district, not the app.
An independent tool. EvalScribe is not affiliated with or endorsed by Kim Marshall or the Marshall Memo. It supports administrators who already use the rubric; if your district has adapted the rubric, tell us at [email protected] and we'll make sure it's reflected.
How EvalScribe works with the Marshall rubric.
A five-step path from the back of the classroom to a finished, standards-aligned evaluation.
Observe the classroom
Formal observation or walkthrough — capture what you see the way you always have.
Enter notes your way
Type, dictate, or snap a photo of your handwritten notes. Smart Scan converts handwriting to text.
Map evidence to criteria
EvalScribe identifies which Marshall criterion each note supports, across all six domains.
Suggested ratings & comments
Get draft ratings and feedback language in the rubric's own voice, across the four-level scale.
Review & finalize
Edit anything you want, finalize the evaluation, and export a clean PDF for your records.
EvalScribe runs natively on iPhone, iPad, and Mac — work where you are.
Capture notes the way you already work.
No retraining required. Whatever your walkthrough style, EvalScribe accepts it.
Snap a photo
Photograph your handwritten notes mid-walkthrough. Smart Scan OCR converts handwriting to text — even messy handwriting, even pedagogical terms.
Dictate as you go
Speak observations directly into the app. Voice dictation captures what you saw without breaking your eye contact with the classroom.
Type directly
Prefer to type? Enter notes straight into the app — they flow into the same DC-aware drafting pipeline as scans and dictation.
Why generic AI falls short for the Marshall rubric.
Pasting your mini-observation notes into a general-purpose chatbot might produce a paragraph that sounds polished — but for Marshall documentation, "sounds polished" isn't the bar. Defensible, criteria-aligned evaluations are.
| Capability | Generic AI (ChatGPT, Claude, etc.) | EvalScribe |
|---|---|---|
| Framework awareness | ✕No concept of the Marshall rubric, its six domains, or its four-level scale. Treats every prompt the same. | ✓Built around the Marshall rubric's actual domains, criteria language, and four-level scale. |
| Criteria mapping | ✕Invents domain and criteria labels that don't match the Marshall rubric. | ✓Maps evidence to the real Marshall criteria using the actual rubric loaded into the app. |
| Mini-observation fit | ✕Built for one big write-up, not the short, frequent observations Marshall is designed around. | ✓Built for fast capture and feedback across many mini-observations — the cadence the rubric expects. |
| The Effective standard | ✕Treats the scale as generic good-to-bad, missing that Effective — not Highly Effective — is the expected level. | ✓Drafts to Marshall's real bar — Effective is solid, expected practice; Highly Effective is truly outstanding. |
| Six-domain breadth | ✕Collapses family outreach and professional responsibilities into vague generalities. | ✓Handles all six domains — including Family & Community Outreach and Professional Responsibilities — individually. |
| Note input | ✕Requires you to retype everything before the AI can do anything with it. | ✓Accepts typed notes, dictation, or photos of handwritten notes via Smart Scan OCR. |
| Feedback language | ✕Vague, generic comments that could describe any teacher. | ✓Professional, evidence-anchored feedback written in the language of the Marshall rubric. |
| Data privacy | ✕Free tiers may use your inputs to train public models. | ✓Enterprise-grade Microsoft Azure. Transient processing. Zero retention. Never used for training. |
| Defensibility | ✕Output you can't trace back to evidence or rubric language. | ✓Every comment is tied to evidence pulled from your notes and the component it supports. |
How much Marshall evaluation time could you reclaim?
Tune the inputs to your reality. The numbers update as you go.
You could reclaim approximately
hours per year
Workdays reclaimed
Total evaluations
Based on early EvalScribe users saving an average of 35 minutes per evaluation. Individual results may vary based on your current workflow.
The best way to see it
is to watch it.
Reading about how EvalScribe drafts a Marshall evaluation isn't the same as watching it happen. The demo walks through a real mini-observation — capture, mapping, scoring, and final export — start to finish.
The draft review screen — every comment, rating, and piece of mapped evidence is fully editable before export.
See an example EvalScribe report.
A real EvalScribe-generated evaluation, formatted exactly as it exports from the app. The workflow is the same across frameworks — useful to share with district stakeholders.
Download sample PDFWhat educational leaders are saying.
Real feedback from administrators across roles — from academic coaches to assistant superintendents.
EvalScribe is a tool that every public school principal should use. This resource literally makes multiple people out of administrators, allowing them to focus their energies on precise observation analysis and quality feedback while leaving EvalScribe to handle the mechanics. This tool will remove virtually all of the drudgery from the classroom evaluation process and would represent an excellent use of school funds.
Dr. Robert Stinson
Principal
After seeing a detailed demonstration of EvalScribe, I came away very impressed with both the concept and the execution. The platform provides administrators with a practical way to quickly capture classroom observations and generate high-quality, customized feedback that remains aligned to their evaluation framework and local expectations. What stood out most was the balance between efficiency and quality. EvalScribe has the potential to reduce administrative workload while helping school leaders provide more timely and meaningful feedback to teachers — a tool with considerable promise for schools and districts.
Glenn Brown
Director of Student Services
Game changer. Absolute game changer for school admin.
Tobin Davidson
Assistant Superintendent · 40+ years in education
This is amazing. I would pay lots of money for this. I’m serious — it’s crazy good. I look forward to using it on a regular basis next year.
Kittie Brown
Assistant Principal
Evaluators spend an inordinate amount of time recording and documenting evaluations. This app is easy to use, affordable, and incredibly efficient. It cuts the time spent recording and documenting evaluations down significantly and frees us up to handle other aspects of the job. Plus, the conference suggestions save on prep time when a conference is warranted.
Angie Brandt
Academic Coach
Built by educators, for educators.
EvalScribe was co-founded by a practicing classroom teacher and his wife — a husband-and-wife team with thirty years of combined classroom and business experience.
Anthony Neely, Ph.D.
Co-founder · Product
A tenured K–12 classroom teacher and Professional Learning Lead. District Teacher of the Year, with nearly 20 years in education and three published books.
Andrea Neely, Ph.D.
Co-founder · Business
Associate Professor of Management. Research and teaching in organizational management.
Your evaluation data stays yours.
Observation notes are processed transiently on Microsoft Azure and are not retained on our servers. Your evaluation records live locally on your device. Notes and evaluations are never used to train public AI models. For districts needing a Data Processing Agreement or FERPA attestation, those are available on request.
Simple, transparent pricing.
Try EvalScribe with three free evaluations. When you're ready to keep going, pick the plan that fits your role.
Individual
For a single evaluator on their own device.
Billed annually per user
- Unlimited evaluations
- All supported frameworks, including the Kim Marshall rubric
- iOS & macOS access
- 3 free evaluations to start
School-wide
For a school with multiple evaluators.
3+ evaluators
- Everything in Individual
- 10% volume discount
- Unique access code issued to your school
- Easy distribution to your evaluators
District-wide
For districts deploying across every evaluator.
Best value for full deployments
- Everything in School-wide
- 20% volume discount
- Single access code for the entire district
- Direct support contact for rollout
All pricing shown is introductory and may change. For school and district plans, EvalScribe issues a unique access code that the buyer distributes to their evaluators.
The Marshall rubric & EvalScribe — common questions.
Does EvalScribe support the Kim Marshall rubric?
Yes. EvalScribe is built around the Kim Marshall Teacher Evaluation Rubric — all six domains, their criteria, and the four-level scale. It maps your mini-observation evidence to the criterion it supports and drafts a rating and feedback in the rubric's own language, from Does Not Meet Standards to Highly Effective.
What are the six Marshall domains?
A. Planning & Preparation for Learning, B. Classroom Management, C. Delivery of Instruction, D. Monitoring, Assessment & Follow-Up, E. Family & Community Outreach, and F. Professional Responsibilities — fifty-four criteria in all across the six domains.
What are mini-observations?
The Marshall rubric is designed to be informed by short, unannounced mini-observations every few weeks, each followed by a quick face-to-face conversation — not a single formal visit. The year-end rubric is filled in from that accumulated evidence. EvalScribe is built to make that observe-often, feedback-fast cadence quick to document.
What are the Marshall rating levels?
Does Not Meet Standards, Improvement Necessary, Effective, and Highly Effective. Effective is the expected level of performance — Marshall is explicit that teachers should feel good about it. Highly Effective is reserved for truly outstanding practice against very demanding criteria.
Is EvalScribe affiliated with Kim Marshall?
No. The rubric was created by Kim Marshall and is published through the Marshall Memo. It is open source and freely adaptable by schools and districts. EvalScribe is an independent drafting tool that supports administrators who already use the rubric; it does not replace your district's official scoring platform.
Does EvalScribe replace evaluator judgment?
No. EvalScribe drafts evidence-mapped evaluations and suggested ratings. The evaluator reviews, edits, and finalizes every document. The professional judgment is always yours — EvalScribe just removes the typing.
Can the generated evaluations be edited?
Yes, fully. Every comment, every rating, and every piece of mapped evidence inside a draft evaluation can be edited before you export the final PDF. You are always the last set of eyes on the document.
Is observation data stored?
Your notes are processed transiently on Microsoft Azure and are not retained on EvalScribe's servers. Your evaluation records sit locally on your device under a local-first architecture. Your data is never used to train public AI models. For more detail, see our Privacy Policy.
How is EvalScribe different from ChatGPT or Claude?
Generic AI tools have no built-in understanding of the Marshall rubric. They invent domains and criteria, ignore the mini-observation cadence it's designed around, hand out Highly Effective freely, and produce vague comments that miss the level distinctions the rubric requires. EvalScribe is built around the Marshall rubric's actual domains and four-level scale — and the mini-observation workflow that surrounds it — from capture through final export.
Where can I read the official Marshall rubric sources?
Kim Marshall publishes the rubric and rationale through the Marshall Memo. The rubric is open source and updated periodically; many districts and state education sites also host copies of the current version.
Which other frameworks are supported?
EvalScribe ships with frameworks from all 50 states and Washington D.C., plus general models such as Danielson, Marzano, Stronge, CEL 5D+, NIET, and Kim Marshall. The complete list is available at evalscribe.com/frameworks.
Don't see your question? Reach out via our contact page — we read every message personally.
Spend less time on paperwork.
Lead your building.
EvalScribe was built to help administrators spend less time documenting Marshall evaluations and more time doing what matters: leading instruction, coaching teachers, and walking buildings.
Three evaluations free — no credit card · Available on iPhone, iPad, and Mac