# Grady

> Grady is an AI-assisted feedback and grading platform custom built for higher education by university faculty. It drafts grades and detailed, question-by-question feedback for student submissions across any subject — from handwritten exams and code notebooks to essays and lab reports — while faculty review, edit, and approve everything before students see it. Grady is privacy-first by design: faculty submissions are never used to train AI models, raw student work is purged after feedback is delivered, and the platform is FERPA, GDPR, and WCAG 2.2 AA compliant with HECVAT 4 available.

## Core information

- [Homepage](https://www.gradyai.com/): Overview of Grady — rigorous AI-assisted grading with faster and richer feedback, faculty in full control, and class-level analytics.
- [How It Works](https://www.gradyai.com/how-it-works): The four-step grading workflow — upload, grade, review, finalize — and an interactive demo.
- [Features](https://www.gradyai.com/features): Accuracy and consistency,detailed student feedback, faculty control, support for every assignment format, every subject, LMS integration, privacy-first architecture.
- [About](https://www.gradyai.com/about): Leadership team, edtech advisors, academic and technical advisory boards, and what Grady stands for.
- [Privacy & Ethics Pledge](https://www.gradyai.com/privacy-pledge): Five core commitments: data anonymization, no data retention, no training on submissions, regulatory compliance, continuous auditing — plus data security and encryption details.
- [Get Started](https://www.gradyai.com/get-started): For faculty and institutions interested in piloting Grady.

## What Grady does

- **Drafts grades and feedback for every assignment format.** Handwritten exams, scratched-out work, hand-drawn graphs and diagrams, typed PDFs, Word documents, code notebooks, and more. No special preparation required.
- **Grades every higher-ed subject.** Computer science, mathematics, chemistry, biology, physics, engineering, economics, marketing, finance, accounting, history, literature, political science, psychology, public health, and others.
- **Builds rigorous, adaptable rubrics.** A proprietary rubric engine interprets faculty instructions and constructs subject-appropriate rubrics for any assessment format.
- **Returns class-level analytics.** Aggregated reports surface common errors, concept gaps, and which students need targeted support.
- **Keeps faculty in full editorial control.** Every grade is reviewable; every comment is editable; nothing reaches students without faculty approval.
- **Saves significant time.** Grading 250 exams typically takes days to weeks by hand; with Grady, the initial pass takes1-2 hours, leaving faculty to edit rather than grade from scratch.

## Leadership

- **Anant Agarwal** — Chief Executive Officer of Grady. Founder and former CEO of edX, the online learning platform launched by MIT and Harvard; former Chief Academic Officer at 2U; Professor of Electrical Engineering and Computer Science at MIT and former director of CSAIL. Recipient of the McGraw Prize for Higher Education, India's Padma Shri, and the Yidan Prize for Education Development.
- **Periklis (Perry) Papakonstantinou** — Co-Founder, Chief Scientific Officer & AI Ethicist. Associate Professor at Rutgers University. Researcher in artificial intelligence and cryptography.
- **Anastasios (Tasos) Sidiropoulos** — Co-Founder & Chief Technology Officer. Associate Professor of Computer Science at the University of Illinois Chicago. Researcher in artificial intelligence and algorithms.

Grady's two faculty founders, Periklis Papakonstantinou and Anastasios Sidiropoulos, are university professors with deep expertise in AI, algorithms, data privacy, cybersecurity, cryptology, and information security. Grady was designed from the start to reflect their commitment to rigorous teaching, faculty agency, and the ethical use of AI in education. They are joined by CEO Anant Agarwal, who founded edX and has spent his career widening access to high-quality education.

## Edtech advisors

- **Paul LeBlanc** — Former President, Southern New Hampshire University; Visiting Scholar, Harvard Graduate School of Education; Co-Founder, Matter and Space.
- **Nina Huntemann** — Senior Consultant at EDT & Partners; former VP of Learning at edX; former Professor at Suffolk University.
- **Mark Haseltine** — Head of Product at Manifest; former CEO of Teachable; former CTO of edX.

## Academic advisory boards

Grady's grading engine has been tested and shaped by faculty across disciplines and institutions, including UC Berkeley, Columbia University, Stanford University, the University of Chicago, The Ohio State University, Rutgers University, and the University of Illinois at Chicago. Advisors span mathematics, computer science, creative writing, history, biological sciences, and accounting, with additional expertise in IEEE-CS publications leadership and academic program direction.

## Compliance, privacy, and security

- **FERPA compliant.** Built to meet the U.S. Family Educational Rights and Privacy Act requirements.
- **GDPR compliant.** Meets European Union data protection standards.
- **WCAG 2.2 AA compliant.** Meets accessibility standards.
- **HECVAT 4 available.** Higher Education Community Vendor Assessment Toolkit documentation provided to institutional procurement and IT teams.
- **No training on submissions.** Grady never trains AI models on student submissions or faculty course material.
- **No data retention beyond display.** Raw student submissions are purged from systems after feedback is delivered.
- **Data anonymization.** A proprietary anonymization engine decouples personally identifiable information from grades and submissions at every stage of processing.
- **Encryption at rest.** AES-256, the encryption standard used by financial institutions and government agencies.
- **Encryption in transit.** TLS 1.3 for every data transfer.
- **Access controls.** Role-based access control and multi-factor authentication.
- **Infrastructure security.** Hosted in SOC 2 Type II certified data centers with continuous monitoring, intrusion detection, and automated threat response.
- **Continuous auditing.** Algorithms are studied on an ongoing basis for transparency, bias, and fairness.

## What Grady is not

- Not a replacement for faculty judgment. Every grade and every piece of feedback is drafted for review; faculty approve everything before students see it.
- Not a tool that learns from student work. Submissions are never used to train AI models.
- Not a black box. Faculty can review, edit, and override every output, and the company publishes its privacy and ethics commitments publicly.

## Contact

- Website: https://www.gradyai.com
- Company: Grady Labs Inc.
- Get started: https://www.gradyai.com/get-started