AI-powered teacher evaluations
for Oklahoma educators.
Cut evaluation documentation from hours to minutes. EvalScribe is built around Oklahoma's TLE Tulsa Model natively — all five domains, all twenty indicators, and the state's five-tier rating scale, in the model's own rubric language.
Saved per evaluation
Built in natively
To capture notes
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.
Oklahoma-specific
The full Tulsa Model built in natively, with its actual indicators and rating scale.
Privacy-first
Transient processing on Microsoft Azure. Never used for AI training.
Built for workflow
Designed around the realities of being an Oklahoma evaluator.
Oklahoma districts choose their TLE framework. EvalScribe is built around the Tulsa Model.
Oklahoma districts select a state-approved qualitative framework under the Teacher and Leader Effectiveness (TLE) system (70 O.S. § 6-101.16). The Tulsa Model — developed by Tulsa Public Schools on a Charlotte Danielson foundation and independently validated by the MET Validation Engine and VARC studies — is the most widely adopted. EvalScribe is built around it natively, with its actual structure and rating scale.
Tulsa Model — Oklahoma TLE
The Tulsa Model is built around five domains and twenty indicators spanning classroom management, instruction, professional growth, interpersonal skills, and leadership. Each indicator is rated on the state's five-tier scale, from Ineffective to Superior. The domains are weighted — Instructional Effectiveness alone carries half the composite — and any indicator scored off Effective carries a documentation requirement: a Personal Development Plan for a 1 or 2, specific supporting comments for a 4 or 5.
Five weighted domains. Twenty indicators.
Classroom Management
6 indicators · 30% weightInstructional Effectiveness
10 indicators · 50% weightProfessional Growth & Continuous Improvement
2 indicators · 10% weightInterpersonal Skills
1 indicator · 5% weightLeadership
1 indicator · 5% weightWeighted composite
A weighted average — not a straight oneThe five-tier rating scale.
Ineffective
Needs Improvement
Effective
Professional standardHighly Effective
Superior
How EvalScribe handles the Tulsa Model
All five domains and all twenty indicators are built into the tool, each with the Tulsa Model's own rubric language across the five levels. EvalScribe maps your evidence to the relevant indicators and drafts ratings and comments with the best-fit logic the model uses — including the narrative comments required whenever a rating lands off Effective (a Personal Development Plan for a 1 or 2, supporting comments for a 4 or 5). The Effective-to-Highly-Effective-to-Superior climb turns on who is doing the intellectual work, and EvalScribe drafts to that distinction rather than defaulting to generic "good/great" language.
Honest scope on the weighting: EvalScribe surfaces each domain's weight (30/50/10/5/5) for reference, but the final weighted composite is completed by the evaluator or your district's system — the app does not compute it.
McREL and Oklahoma TAP — coming in a future build.
Oklahoma's TLE system approves several qualitative frameworks. EvalScribe natively supports the Tulsa Model and the Marzano framework today. Support for McREL and the Oklahoma TAP Teaching Standards framework is on the development roadmap and coming in a future update. Oklahoma evaluators using EvalScribe today won't be left behind as additional framework support is added.
How EvalScribe works with the Tulsa Model.
A five-step path from the back of the classroom to a finished, framework-aligned evaluation.
Observe the classroom
Walkthrough or formal observation — 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 indicators
EvalScribe identifies which of the 20 Tulsa indicators each note supports, across all five domains.
Suggested ratings & comments
Get draft performance-level ratings and feedback language in the Tulsa Model's own voice — including the narrative comments required off Effective.
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 Tulsa-aware drafting pipeline as scans and dictation.
Why generic AI falls short for the Tulsa Model.
Pasting your walkthrough notes into a general-purpose chatbot might produce a paragraph that sounds polished — but for Tulsa Model documentation, "sounds polished" isn't the bar. Defensible, framework-aligned evaluations are.
| Capability | Generic AI (ChatGPT, Claude, etc.) | EvalScribe |
|---|---|---|
| Framework awareness | ✕No concept of the Tulsa Model, its 20 indicators, or its five-tier scale. Treats every prompt the same. | ✓Built around the Tulsa Model's actual structure, rubric language, and five-tier rating scale. |
| Indicator mapping | ✕Hallucinates indicator numbers and invents rubric language that doesn't exist in the model. | ✓Maps evidence to the real 20 Tulsa indicators using the actual rubric loaded into the app. |
| Rubric scoring | ✕Inconsistent ratings between teachers — and between sessions on the same teacher. | ✓Suggests best-fit performance ratings across the five tiers based on captured evidence. |
| Level distinctions | ✕Defaults to generic "good/great" framing that misses the Effective-to-Highly-Effective distinction the model requires. | ✓Rubric-aware scoring that turns on who's doing the intellectual work, the way the Tulsa Model intends. |
| Domain weighting | ✕No awareness the domains are weighted at all; treats every indicator as if it counts the same. | ✓Surfaces each domain's weight (30/50/10/5/5) for reference. The final weighted composite is completed by the evaluator or district system. |
| Narrative requirement | ✕No awareness that any rating off Effective needs written justification. | ✓Drafts the required supporting comments and Personal Development Plan language for off-Effective ratings. |
| 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. |
| 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 indicator it supports. |
How much Oklahoma 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 Tulsa Model evaluation isn't the same as watching it happen. The demo walks through a real 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 Tulsa Model
- 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.
Oklahoma TLE & EvalScribe — common questions.
Does EvalScribe support the full Tulsa Model?
Yes — all five domains and all twenty indicators are built in natively, each with the Tulsa Model's own rubric language across the five performance levels, plus the look-fors and the distinctions that separate adjacent ratings. EvalScribe maps your evidence to the right indicators and drafts ratings and comments in the model's voice.
Does EvalScribe calculate the weighted composite score?
No — and we're deliberately clear about that. The Tulsa Model's composite is a weighted average (Classroom Management 30%, Instructional Effectiveness 50%, Professional Growth & Continuous Improvement 10%, Interpersonal Skills 5%, Leadership 5%). EvalScribe drafts the indicator-level ratings and comments and shows each domain's weight for reference, but the final weighted composite is completed by the evaluator or your district's system, not the app.
Does EvalScribe support Marzano, McREL, or Oklahoma TAP?
Marzano, yes — EvalScribe natively supports the Marzano framework as well as the Tulsa Model. McREL and the Oklahoma TAP Teaching Standards framework are not yet supported; dedicated support for those is on the development roadmap and coming in a future update.
What about the PL Focus?
OSDE's Professional Learning Focus is the educator's own professional-learning component and sits outside the observation rubric. EvalScribe handles the observation-based Tulsa Model evaluation — the write-up, the indicator-level ratings, and the feedback language. The PL Focus stays teacher-authored, which is exactly where the framework intends it.
Does EvalScribe replace evaluator judgment?
No. EvalScribe drafts evidence-mapped evaluations and suggested performance-level ratings. The certified evaluator reviews, edits, and finalizes every document. The professional judgment is always yours — EvalScribe just removes the typing.
Does EvalScribe handle the quantitative or student-growth component?
No. EvalScribe addresses the qualitative observation component of the Tulsa Model. Any student-growth or quantitative measures your district applies are handled outside the app.
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 Tulsa Model. They invent indicators that don't exist, score inconsistently, have no concept of the domain weighting, and produce vague comments that miss the level distinctions the model actually requires. EvalScribe is built around the Tulsa Model's actual structure and rating scale — and the evaluator workflow that surrounds it — from capture through final export.
Is EvalScribe available for iOS and macOS?
Yes. Individual evaluators can download EvalScribe from the Apple App Store on both iOS and macOS, with a free trial that includes three evaluations.
Can schools or districts purchase licenses for multiple evaluators?
Yes. School-wide licensing starts at three or more evaluators ($90/user/year), and district-wide licensing is $80/admin/year. When you order, EvalScribe issues a unique access code that you distribute to your evaluators — no separate provisioning required. Email [email protected] to get started.
Where can I read the official TLE and Tulsa Model sources?
The primary sources are maintained by OSDE and Tulsa Public Schools: the OSDE Teacher and Leader Effectiveness (TLE) pages and the TLE qualitative components library, which houses the Tulsa Model Observation and Evaluation Handbook. Statutory authority is 70 O.S. § 6-101.16 and Okla. Admin. Code § 210:20-41-1.
Which other state 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 Oklahoma evaluators spend less time documenting 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