# audio spectrum

audio spectrum is a remote worker on Himalayas, PFE Intern at TELNET, based in Tunisia.

Source: https://himalayas.app/@audiospectrum

```text
I build AI-driven software for cybersecurity, computer vision, and LLM applications.
```

| Detail | Value |
| --- | --- |
| Availability | Looking for a job |
| Location | Tunisia |
| Authorized to work in | Tunisia |
| Job categories | Artificial Intelligence Engineer, Machine Learning Engineer, Software Engineer, Cybersecurity Engineer, Computer Vision Engineer |
| Skills | AI, Python, Java, C, C++, PHP, Spring Boot, Angular, MySQL, PostgreSQL, TensorFlow, PyTorch, Pandas, Keras, Large Language Models, Lang Chain, Prompt Engineering, LangGraph, scikit learn, NumPy, OpenCV, Hugging Face Transformers, SHAP, LIME, Adversarial Machine Learning, Docker, Git, Graph RAG, RAG, Random Forest, Extra Trees, XGBoost, LightGBM, YOLO, Flask, PyQt, Object Oriented Programming, Model View Controller, Computer Vision, Android Malware Detection, Cybersecurity, data preprocessing, Model Optimization, Dataset Creation |
| LinkedIn | https://linkedin.com/in/abderrahmen-marrakchi |

## What I'm looking for

```text
I'm looking to contribute to innovative software and AI solutions, especially work involving AI-driven applications, machine learning, cybersecurity, computer vision, and LLM-based systems.
```

## About

```text
I've built AI-driven applications spanning Android cybersecurity, computer vision, and LLM-based software. At TELNET, I developed ATT&CKDroid, an explainable Android malware detection application that maps permissions and sensitive APIs to MITRE ATT&CK Mobile techniques.

I trained an ensemble of Random Forest, Extra Trees, XGBoost, and LightGBM models to reach 91.37% accuracy, with GPT-4o-mini escalation increasing accuracy to 92%. At ENIS–Moncton, I evaluated Android malware models against adversarial attacks on Compute Canada infrastructure and developed defenses that reduced attack success by about 50%.

I've also built a YOLO, Flask, and Android bakery detection system, a PyQt5 tool for automated DLT filtering, and RAG pipelines for LLM test-oracle generation and arXiv paper recommendations.
```

## Experience

### PFE Intern

TELNET. Feb 2026 - Jun 2026

```text
Developed ATT&CKDroid, an AI-powered Android malware detection app using a GraphRAG-based pipeline. Trained an ensemble model achieving 91.37% accuracy, escalating complex cases to GPT-4o-mini for deep analysis and pushing accuracy to 92%.
```

Skills: Graph-RAG, Android, Machine-Learning, Random-Forest, Extra-Trees, XGBoost, LightGBM, GPT-4o-Mini, Cybersecurity

### Python Development Intern

KPIT. Jul 2025 - Aug 2025

```text
Designed a PyQt5 desktop application that converts DLT Viewer XML filters into executable Python scripts for automated DLT filtering. Architected the application using OOP and MVC to ensure long-term scalability and maintainability.
```

Skills: Python, PyQt, Object-Oriented-Programming, MVC, DLT

### Adversarial Machine Learning Research

ENIS – Moncton, Canada. Nov 2024 - Apr 2025

```text
Reproduced state-of-the-art Android malware detection models and ran large-scale experiments on Compute Canada HPC infrastructure. Applied adversarial attack techniques and designed defense strategies, reducing attack success rate by ~50%.
```

Skills: Adversarial-Machine-Learning, Android, Malware-Detection, HPC, APK-Analysis

### Computer Vision Intern

CRNS. Jun 2024 - Jul 2024

```text
Built a real-time bakery product detection and counting system using YOLO, Flask, and an Android application. Enhanced recognition performance by integrating CLIP into the detection pipeline, achieving 80% accuracy and a 15% improvement in recognition performance.
```

Skills: Computer-Vision, YOLO, Flask, Android, CLIP

## Education

### National School of Engineers of Sfax (ENIS)

Engineering Cycle, Computer Engineering and Applied Mathematics. 2023 - Present

```text
Engineering Cycle in Computer Engineering and Applied Mathematics, final year.
```

### Preparatory Institute for Engineering Studies of Sfax (IPEIS)

Preparatory Cycle, Mathematics and Physics. 2021 - 2023

```text
Preparatory Cycle in Mathematics and Physics.
```

### Farhat Hached Pioneer High School

Baccalaureate, Experimental Sciences. 2020 - 2021. Grade: Highest Honors

```text
Baccalaureate in Experimental Sciences, passed with highest honors.
```

### NVIDIA

Building LLM Applications With Prompt Engineering, Prompt Engineering

```text
Certification in Building LLM Applications with Prompt Engineering.
```

### NVIDIA

Fundamentals of Deep Learning, Deep Learning

```text
Certification in Fundamentals of Deep Learning.
```

## Tech stack

- PyQt
- NumPy
- Pandas
- MySQL
- PostgreSQL
- Spring Boot
- OpenCV
- Python
- Java
- PHP
- C++
- C
- TensorFlow
- PyTorch
- scikit-learn
- Keras
- Flask
- Git
- Docker
- XGBoost
- LightGBM
- LangChain
- LangGraph
- GraphRAG

To contact audio, message them from this profile with a Himalayas employer account: [Hire on Himalayas](https://himalayas.app/hire).

## People also viewed

- [Ahlem Ammar](https://himalayas.app/@ahlemammar1): Data Scientist, VIVIAClinical Network
- [Haythem Chaabeni](https://himalayas.app/@haythemchaabeni): AI & LLM Intern, Audasoft Technology
- [Nada Arfaoui](https://himalayas.app/@nadaarfaoui): Machine Learning & Cloud Engineering Intern, Infor
- [jihed Ayari](https://himalayas.app/@jihedayari): AI Engineering Intern, Smartovate Technologies
- [Ameni Krichen](https://himalayas.app/@amenikrichen): Co Organiser & Speaker, TwinTech Talks
- [Oussema Benameur](https://himalayas.app/@oussemabenameur): AI Lead, Electronica Club ENISo
- [Meriem Kilani](https://himalayas.app/@meriemkilani): ML Engineer – Services Team, Multiverse Computing
- [elhaddad oussama](https://himalayas.app/@elhaddadoussama): Gait Analysis – MITACS GRI, University of New Brunswick
- [Oumaima Souguir](https://himalayas.app/@oumaimasouguir): AI & Machine Learning Developer, Omdena
- [Youssef Ouertani](https://himalayas.app/@youssefouertani): Freelance AI & Full Stack Engineer, Independent