README.md

metadata

title: Abyssinia Intelligence v15
emoji: 🧬
colorFrom: blue
colorTo: gray
sdk: docker
pinned: false
app_port: 7860
app_file: app.py

Abyssinia Intelligence v15

Clinical AI Architect: Dawit

A high-fidelity clinical decision support system (CDSS) for molecular oncology surveillance and toxicity forecasting. This platform utilizes deterministic pharmacological mapping to provide therapeutic guidance based on genomic alterations.

🛠 Technical Stack

• Frontend: Vite + React 18 (TypeScript)
• State Management: Zustand with persist middleware (Local Vault)
• Styling: Tailwind CSS (Gemini-Slate aesthetic, 0% Emojis)
• Backend/Orchestration: Flask + Gunicorn (Dockerized)
• Cryptography: Web Crypto API (SHA-256 Digital Signatures)

🏗 Architectural Layers

1. Molecular Analysis Engine

Maps genetic mutations (e.g., KRAS G12C, EGFR L858R) to NCCN-aligned therapeutic protocols and first-line interventions.

2. Toxicity Forecasting Model

Calculates Adverse Event (AE) risk scores using a baseline risk $R_b = 0.05$ and mutation-specific adjustments.

• Formula: $Risk = \text{Baseline} + \text{Mutation Weight} + \text{Pharmacological Interaction}$

3. The Clinical Vault (Security)

All session data is signed using a SHA-256 hash. This "Digital Seal" ensures that any exported report can be verified against the local state to prevent data tampering in clinical settings.

🚀 Local Setup

# Install dependencies
npm install

# Initialize Clinical Vault paths
npm run init-vault

# Start Development Server
npm run dev