// cat case-file.md
BioScan
🏆 1ST PRIZE · CLOUD & SECURITY · BAL DES PROJETS 2026
A secure medical document platform built end-to-end with team SKeySec at ESPRIT — from a hardened, segmented network in GNS3 up to an ML-powered web application.
role: Scrum Master
team: SKeySec (5)
context: Bal des Projets 2026, ESPRIT
github.com/Givemeboga/BioScan ↗
7
segmented VLANs
4
role-based dashboards
5
person team
20+
REST domains, Swagger-documented
// network --infrastructure
A dual pfSense firewall architecture segments the platform into seven VLANs, with defense-in-depth between every zone. Core services run on dedicated segments; everything is observable through a central SIEM stack — designed, built and validated entirely in GNS3.
VLAN SEGMENTS
VLAN 10 · DMZ
VLAN 20 · Database
VLAN 30 · Médecin
VLAN 50 · Technicien
VLAN 60 · Backend / AI
VLAN 99 · Administration
Stockage
Dual pfSense 2.7
Perimeter + internal firewalling, NAT and per-VLAN ACL rules
BIND9 + Samba AD
Internal DNS in the DMZ and Active Directory identity services
Grafana + Loki SIEM
Logs centralized from pfSense, web, database and VPN for real-time supervision
// webapp --security
The application layer handles medical document intake with a security-first pipeline: every upload is scanned, extracted, analyzed and reported automatically.
FastAPI + React 18 + PostgreSQL
Role-based platform for Admin, Doctor, Technician and Patient
Analysis engine
Ingests CSV, Excel, PDF, JSON & XML bilans; flags anomalies against clinical reference ranges with Pandas + Scikit-learn
JWT + 2FA / AES-256
TOTP two-factor auth, password-reset flow, encrypted sensitive data
ClamAV
Malware scanning on every uploaded document
Tesseract OCR
Automated anonymization of patient documents
ReportLab
Automated PDF & JSON report generation
Admin console
User/doctor/technician CRUD, audit-log security feed, analytics dashboard
// app --patient-portal
The patient-facing landing brings a clinical bilan into plain language — validated blood-work status, trend charts for cholesterol, glucose and white cells, the next appointment and a direct line to the medical team.
