Process Summary
First Draft
The Process
Document Your Process
Revision
Draft-by-draft Changes
Landing Page
Start Building

Proof of Process

The Problem: A Crisis in Classrooms Today

AI writing detectors in schools are unreliable and biased. A 2023 Stanford study found that GPT detectors correctly identified essays written by native English speakers almost every time, but flagged more than half of essays written by non-native English speakers as AI generated.

Real cases confirm this pattern. One professor used Turnitin's AI detector and flagged over 90 percent of a student's paper as AI written. The student had to sit through a meeting defending work they actually wrote themselves. This is happening in classrooms today, punishing the students with the least power to push back.

Who experiences this? Multilingual and non-native English speakers face the most harm. They are disproportionately misflagged by detectors, creating a system that claims to protect academic integrity but actually reinforces existing inequities.

Why it matters? Students lose trust in the educational process. Teachers are forced to navigate false accusations. The burden falls on the accused to prove a negative, which is inherently unfair. Current solutions add more surveillance layers instead of empowering students to demonstrate their authenticity.

The Solution: Flip the Problem

Instead of trying to prove a student's writing is NOT AI generated, give students an easy way to show that it IS genuinely theirs.

Proof of Process lets a student paste in multiple versions of a piece of writing: a rough draft, a revision, a final version, along with optional notes about why they made certain changes. The tool shows a visual diff between the drafts so you can actually see the writing take shape. An AI model then generates a short, neutral summary describing how the piece evolved across those drafts.

The output is a shareable portfolio a student can attach to their work as honest evidence of their process.

What makes this different? It shifts from accusation-based evaluation to evidence-based verification. It empowers students instead of surveilling them. It replaces hope with proof.

How I Built It

The frontend: a clean, professional interface with three main screens. A landing page explaining the concept. A form for entering draft versions and reflections. A results page showing the visual diff alongside the AI-generated summary. The design uses a blue and teal palette to feel trustworthy and academic.

The backend: a small REST API with a single endpoint that accepts draft versions and reflection, sends them to an LLM, and returns a generated process summary as JSON. I kept it intentionally minimal since this is a proof of concept. No user accounts. No database of past submissions. Just a straightforward request and response flow.

The critical work was the AI prompt itself. Early versions of the summary sounded judgmental, almost like grading the student's honesty. I rewrote the prompt several times until the tone stayed neutral and descriptive rather than evaluative. Using AI responsibly and for genuine social good was a priority from the start.

Challenges and Learning

Getting the frontend and backend talking locally took longer than expected, mostly around cross-origin configuration. But the bigger challenge was thinking carefully about how to use AI in a way that actually served students, not policed them.

I learned that the most convincing AI-powered feature is not always the flashiest. My AI call does a single, focused task: summarizing how writing changed. Keeping it narrow made it easier to get right.

I also learned the research behind AI detection bias. Instead of a vague pitch about AI being scary, I could point to a specific 2023 Stanford study and explain exactly how the tool addresses its failure modes. This changed how I talk about the project entirely.

The Impact

This creates real value for students and educators:

For students: You no longer defend yourself against an accusation. You show your work instead. You regain agency. You have a fair way to prove your authenticity.

For teachers and academic integrity offices: You have a fairer way to evaluate authorship. You move from suspicion-based conversations to evidence-based ones.

For multilingual and non-native English speakers: The group most harmed by biased detectors now has a tool built specifically to protect them.

Proof of Process turns a stressful, often biased accusation process into a simple act of showing your work, which is something every student is already capable of doing.

What's Next

If I kept building, I would add PDF export for proper portfolio generation. A lightweight view for teachers to see class portfolios in one place. Real-world testing with students and teachers to see whether a process summary actually changes how conversations about academic integrity happen in practice. And partnerships with schools serving multilingual student populations, the group most affected by the bias this project addresses.