Legal & Compliance

Introduction

This Privacy Policy describes how the Proprietary AI Grading Engine (the "Platform") collects, uses, and protects information. We provide automated assessment tools for educators, prioritizing the security of student work and the privacy of the classroom.

Information Collection & Use

• Educator Data: We collect names, institutional emails, and school affiliations to manage accounts and provide service updates.
• Student Data: The Platform processes student names, IDs, and images/PDFs of handwritten math scripts solely for the purpose of automated assessment as directed by the Data Controller.
• Proprietary Assessment Logic: Digitized scripts are analyzed via our Proprietary AI Grading Engine. This process is strictly "stateless"—data is processed for real-time inference and is not retained by sub-processors for any secondary purpose.

Data Protection & Sharing

• Zero-Training Guarantee: We utilize Advanced Computational Models via secure APIs. We maintain strict contractual agreements ensuring that student data is never used to train, retune, or improve the foundation models of any Third-party AI Sub-processors.
• No Commercialization: We do not sell, rent, or monetize student or teacher data for advertising or any other commercial purpose.

1. Acceptance of Terms

By accessing the Platform, Schools, Districts, and individual Educators ("Users") agree to these Terms. The Platform is a tool to supplement—not replace—professional educator judgment.

2. Proprietary Rights

All logic, algorithms, and software used in the Proprietary AI Grading Engine are the exclusive intellectual property of the Platform. Users are granted a non-exclusive license to use the Platform for educational assessment.

3. AI Feedback Disclaimer & Human-in-the-Loop

• Supplement, Not Substitute: The Platform provides automated suggestions based on assignment rubrics. The Teacher of Record remains the final authority on all grades.
• Accuracy: While our Advanced Computational Models are highly accurate in interpreting mathematical notation, the Educator is responsible for reviewing AI-generated feedback before finalizing student records.

I. Data Ownership and Control

The School or District (the "Data Controller") retains all rights, title, and interest in and to Student Data. The Platform (the "Data Processor") acts as a "School Official" under FERPA with a legitimate educational interest.

II. Technical Safeguards

• Data in Transit: All data sent from the User to the Platform is secured using industry-standard TLS 1.3 encryption.
• Data at Rest: PII and student scripts are stored in a secure cloud database infrastructure using AES-256 encryption.
• Stateless AI Processing: Student PII sent to Third-party AI Sub-processors for math analysis is processed in a stateless environment. No data is stored or cached by the sub-processor beyond the duration of the inference request.

Architecture Overview

The Platform utilizes a secure, cloud-hosted architecture optimized for high-integrity educational assessment. Our stack leverages a NoSQL database for flexible data handling and real-time backend services for rapid grading turnaround.

The "Zero-Training" Inference Pipeline

A primary concern in AI governance is the "leakage" of PII into public AI models. Our architecture solves this through:

• Isolated Inference: We utilize Enterprise-grade Advanced Computational Models through private API endpoints.
• Explicit Opt-Out: Our enterprise-tier agreements with sub-processors explicitly prohibit the use of transmitted data for "Global Model Training."
• Data Minimization: We only transmit the portions of the math script necessary for assessment.

Security Protocols