blockchain-edu

Inspiration

Campus-industry collaboration gives students valuable real-world experience, but project evidence is often scattered across documents, spreadsheets, and chat tools. This makes it difficult for schools, teachers, company mentors, and students to verify what was submitted, when it was reviewed, and how the final evaluation was calculated.

Blockchain Edu was created to make the entire process transparent, traceable, and easier to manage while protecting students' personal information.

What it does

Blockchain Edu is a role-based project management and student evaluation system designed for four participants:

School administrators
Faculty advisors
Company mentors
Students

The system provides:

Project publishing and status management
Student practice profiles with privacy masking
Submission evidence registration
SHA-256 integrity verification and tamper detection
Evaluations from faculty and company mentors
AHP-entropy combined evaluation
Traceable evaluation results
Audit logs and test reports
JSON data export

The current prototype uses a simulated blockchain ledger. It generates cryptographic digests and transaction-like identifiers in the browser, allowing the complete evidence verification process to be demonstrated without requiring a blockchain node or wallet.

How we built it

The prototype was built with HTML5, CSS3, and vanilla JavaScript. It runs directly in a browser without requiring installation or a backend server.

The Web Crypto API generates SHA-256 digests for submitted evidence. Browser localStorage stores the demonstration data. The evaluation module combines AHP expert weights with entropy-based data weights and displays the consistency ratio, dimension scores, combined weights, and final result.

The system also implements role-based permissions so that sensitive student information is only visible to authorized users.

How we used Codex

OpenAI Codex served as the primary engineering partner throughout the project.

We used Codex to:

Transform the original research concept into a working system architecture
Design the role and project state models
Implement the evidence hashing and verification workflow
Build the evaluation, traceability, and audit modules
Improve the user interface and responsive layout
Check the JavaScript implementation
Write project documentation
Publish the source code to GitHub

Codex helped us move from a thesis concept to a coherent and testable prototype during OpenAI Build Week.

Challenges we ran into

One major challenge was balancing traceability with privacy. Evaluation results need enough evidence to be audited, while students' personal information must remain protected.

Another challenge was demonstrating blockchain-style evidence honestly without claiming that the prototype was already connected to a production blockchain. We therefore clearly implemented the current ledger as a local simulation that can later be replaced by a smart contract or consortium-chain adapter.

Designing an explainable evaluation process was also challenging. Instead of returning only a final score, the system exposes its weights, consistency checks, dimension scores, source evaluations, and related evidence.

Accomplishments that we're proud of

Implemented a complete workflow covering four campus roles
Created a live demonstration of evidence hashing and tamper detection
Built an explainable multi-stakeholder evaluation algorithm
Connected final scores with materials, project stages, and audit records
Added privacy-aware access to student information
Delivered a self-contained prototype that runs immediately in a browser

What we learned

We learned that trust in educational assessment requires more than immutable records. It also depends on explainable algorithms, clear permissions, privacy protection, and an interface that allows non-technical users to understand how a result was produced.

We also learned how Codex can act as an engineering collaborator across architecture, implementation, testing, documentation, and delivery.

What's next

The next version will introduce an OpenAI-powered evidence assistant that can summarize submissions, identify missing evidence, and draft review suggestions while keeping human mentors responsible for final evaluations.

We also plan to add: