Mohamed Zine El Abidine Kacem

Core · summary

Highly analytical Backend Developer and AI Engineer with 4+ years of experience specializing in scalable web applications and AI systems. Expert in Python/Django architecture, performance optimization, and deep learning for medical data (ECG). Focused on secure, high-performance systems.

Thinking process

Constraint-first engineering — architecture before implementation.

01
Analyze constraints
Map requirements, risks, and non-negotiables before touching code.
02
Design system architecture
Define boundaries, contracts, and failure modes across services.
03
Optimize performance
Profile queries, queues, and hot paths; measure before and after.
04
Ensure scalability & security
Design for load, isolation, and least-privilege from day one.

System modules

Skills as loadable subsystems.

REST contracts, validation, versioning, and predictable error surfaces.

Django
DRF
FastAPI

Experience

Problem → solution → result.

FlexUp

2023–Present · Paris

Lead Backend Developer

Problem: Backend complexity and data access patterns risked slowing product delivery.
Solution: Owned architecture decisions, tightened PostgreSQL/MySQL usage, and grew reliable background processing.
Result: Stronger query discipline, resilient async work, higher test coverage, and calmer production operations.

· Led backend architecture decisions
· Optimized PostgreSQL/MySQL queries
· Built background processing systems
· Improved testing coverage and system stability

NEXEO

2025–2026 · Remote

Backend Developer

Problem: NFC digital card flows needed secure, high-concurrency APIs without brittle auth.
Solution: Implemented hardened REST APIs with explicit permission layers and performance-minded endpoints.
Result: Secure issuance flows, clearer auth boundaries, and systems tuned for concurrent traffic.

· Built NFC digital card backend systems
· Designed secure REST APIs
· Implemented authentication & permission systems
· Ensured high concurrency performance

KIO

2023–2024 · Algeria

Backend Developer

Problem: Store creation was slow; feature work needed safer data access patterns.
Solution: Reworked processing paths, added an email validation security layer, and engineered precise SQL for new capabilities.
Result: 40% faster store creation, stronger validation posture, and features backed by sound queries.

· Reduced store creation processing time by 40%
· Built email validation security layer
· Designed complex SQL queries for new features

Projects · system view

Cardiac Arrhythmia Detection Software

Pipeline topology — hover nodes.

ECG Data → Preprocessing → Neural Network → Prediction

— Deep learning with TensorFlow/Keras
— ECG signal analysis
— Full pipeline: preprocessing → training → deployment

How I build systems

Data flow, decisions, and scaling levers — interactive overview.

API Layer

Thin controllers, explicit schemas, observable failures.

Workers (Celery)

Idempotent tasks, dead-letter strategy, concurrency caps.

Database

Migrations with roll-forward plans; indexing as a feature.

Cache

TTL discipline; cache aside; invalidate on writes.

→ Horizontal scaling of stateless APIs
→ Read/write path separation where it pays off
→ Defense in depth for auth and permissions

Performance

40%

store creation improvement (KIO)

Concurrency

High

throughput-oriented systems

APIs

Production

hardened REST surfaces

Reliability

Testing

coverage & stability focus

Contact

Let's build scalable systems.

Doha, Qatar

GitHub LinkedIn Email