MemoryVerse AI

Your academic story, connected.

MemoryVerse AI is a Digital Identity System built for the MemoryVerse AI '26 Hackathon. It transforms scattered student evidence—certificates, projects, internships, resumes, achievements, and learning records—into a structured, searchable digital journey.

The problem

Students accumulate valuable documents across drives, mailboxes, and devices. Traditional storage can save those files, but it cannot explain how a certification connects to a skill, project, internship, or career path.

MemoryVerse turns those disconnected records into a searchable personal knowledge repository—while keeping the original source available.

Features

• Memory library: upload and browse academic or professional evidence.
• Intelligent organization: local categorization into projects, skills, certifications, internships, achievements, and academic records.
• Relationship engine: connects documents through shared extracted skills and shows evidence confidence.
• Journey timeline: converts dated milestones into a visual growth narrative.
• Natural-language retrieval: ask questions such as “Show my AI projects” or “What supports my Python skill?”
• Original evidence access: open the source record from every relevant result.
• Free and local: no account, API key, backend, or cloud dependency is required for the demo.

Demo walkthrough

1. Open the app and ask “Show my AI projects.”
2. Ask “What supports my Python skill?” to retrieve related evidence.
3. Open Connections to show why the system links certificates, projects, and internships.
4. Open Journey timeline to show growth from 2023 to 2026.
5. Open Memory library, upload a document, and show the new local record.
6. Click Open original to demonstrate source preservation.

Run locally

No installation is needed.

# Option 1: open index.html directly in a browser

# Option 2: run the full local Python application
python server.py

Then visit http://localhost:8080. This starts the Python backend for persistent uploads, local analysis, retrieval, and original-file downloads.

Architecture

flowchart LR
  A[Student documents] --> B[Local ingestion]
  B --> C[Category and skill extraction]
  C --> D[Structured memory records]
  D --> E[Relationship engine]
  D --> F[Journey timeline]
  G[Natural-language question] --> H[Weighted retrieval]
  E --> H
  H --> I[Evidence-backed result]
  I --> J[Original source]

Loading

The current prototype uses transparent, browser-side keyword and weighted matching so it can run for free and offline. The production-ready AI evolution is documented in docs/ARCHITECTURE.md.

Tech stack

• HTML5
• CSS3
• Vanilla JavaScript
• Browser localStorage for local metadata persistence
• Browser Object URLs for original-file access in the active session

Project structure

memoryverse-ai/
├── index.html                 # App structure and screens
├── styles.css                 # Responsive visual design
├── app.js                     # Ingestion, retrieval, links, timeline
├── server.py                  # Python API: upload, analysis, search, storage
├── docs/
│   ├── ARCHITECTURE.md         # Current and production AI architecture
│   ├── architecture-diagram.svg# Upload-ready architecture diagram
│   └── THOUGHT_PROCESS.md      # Design decisions for evaluation
├── LICENSE
└── README.md

Evaluation alignment

Hackathon criterion	MemoryVerse response
Quality of AI organization	Auto-categorization, extracted skill tags, searchable evidence
AI/ML technical stack	Clear production plan for NLP, embeddings, vector search, RAG, and relationship mapping
Innovation & UX	A personal growth timeline with explainable relationships
Clarity & architecture	Visible data flow, source-backed retrieval, and documented decisions

Documentation

• Architecture and AI evolution
• Thought process and design decisions

License

Released under the MIT License.