Skip past primary navigation

How Quill Builds Curriculum

Human-Written Texts AI-Powered Feedback

See how we research, write, and test every Quill Reading for Evidence activity.

The Common Assumptions

What Does "Human-Written" Really Mean?

When schools hear "AI-powered," they often assume the texts are too. Here's what's actually happening.

Quill Our Competitors
Every text is written by a curriculum expert and reviewed by a team of editors and fact-checkers. "AI-powered" means the texts are AI-written.
Each activity is designed around a specific learning outcome. Activities are generated from a broad topic prompt.
Feedback is hand-built per prompt, then refined through extensive review and revision. Feedback is generic AI output.
Quality depends on the humans who build, test, and validate every activity. Quality depends on the model.

The Process

How We Build Reading for Evidence Activities

Six steps. Every one of them starts with a human.

  1. Step 1

    Define the learning goals

    Aligned to state and national standards and to our curriculum partners.

  2. Step 2

    Design the writing task

    Build a writing prompt and define exemplars and success criteria.

  3. Step 3

    Draft and revise the text

    Draft an original nonfiction text. Each text then receives a series of reviews with both internal and external editors.

  4. Step 4

    Train AI with teacher-written feedback

    Write feedback on 50–100 anticipated student responses per prompt, then train the AI to replicate that expert coaching.

  5. Step 5

    Test with teachers

    Conduct testing with our Teacher Advisory Council and refine the feedback based on thousands of student responses.

  6. Step 6

    Review and refine

    Read through hundreds of student responses from the beta, then make final refinements before public release.

"My school was originally using [competitor] for independent ELA practice, but I found [it] unhelpful, specifically with regard to its lack of high-quality writing activities… Reading for Evidence activities have a variety of high-quality prompts, and, even more importantly, Quill provides immediate feedback to the students' work.

These activities and the immediate feedback have been a great help for developing my students' reading and writing skills by both reducing the amount of time I have to spend creating writing prompts and giving feedback."

J. Geisler, Quill Teacher Advisory Council

Our Line on AI

Where AI Fits, and Where it Doesn't

Every "AI-powered" company draws this line somewhere. Here's where we draw ours.

What AI is used for

  • Coach millions of students with real-time feedback as they revise their own writing
  • Spot patterns in anonymized student responses to validate the response types predicted by curriculum developers.

What AI is not used for

  • Write or level the texts students read
  • Define learning goals
  • Design writing tasks or define what a strong response looks like

Partnerships

Built in Partnership with Mission-Aligned Organizations

Mission-aligned curriculum partners help us support the content areas teachers ask for most.

Social Studies

OER Project

Activities aligned to OER Project units, so writing practice reinforces the history students are reading.

View Social Studies Activities

AI Literacy

aiEDU

Writing activities that build AI literacy as students weigh AI's promises and pitfalls through evidence-based arguments.

View AI Literacy Activities

Middle School ELA

Fishtank Learning

Literature-based activities tied to the novels students are reading in Fishtank's middle school ELA units.

View Literature Activities

Go Deeper

Learn More

Funders & Partners

Read our Ethical AI Playbook

How we think about human-AI collaboration in K-12 literacy.

Download the playbook

District & Curriculum Leaders

Watch Quill at Work

A walkthrough with our team — across writing and reading, and how districts are rolling it out.

Get a demo

Try a Reading for Evidence Activity with Your Class — Free

Get started