Ayoub MLAOUAH

About Me

Turning Data into Decisions

I'm a Lead Data Scientist at DataPi, where I collaborate with Asten to support retail stores in fraud detection and performance tracking. My work focuses on using advanced data analysis and machine learning techniques to detect anomalies in transactions, identify behavioral patterns, and provide actionable insights.

At DataPi, we've developed interactive dashboards with Metabase for reporting and decision-making, as well as IRIS — an intelligent multi-agent AI assistant connected to the database, enabling real-time and smart interaction with business data via Telegram.

I also designed and deployed secure on-premise AI infrastructure for BIM workflows at AuxiBIM, building Ray-based distributed GPU environments and integrating Ollama with Llama/DeepSeek to deliver privacy-preserving LLMs.

🤖

AI & ML Engineering

LLMs, Multi-Agent Systems, Computer Vision

📊

Data Analytics

Fraud Detection, Pricing Intelligence, BI

☁️

Cloud & Infrastructure

Azure, Docker Swarm, Ray GPU, On-Prem

📚

Research & Teaching

BCI, Deep Learning, NVIDIA DLI Instructor

Career Path

Professional Experience

A journey through AI, Data Science, and Software Engineering.

Aug 2025 — Present

Consultant — Lead Data Scientist & AI Engineer

AuxiBIM

Djerba, Tunisia

Designed and deployed secure on-premise AI infrastructure for BIM workflows. Built Ray-based distributed GPU environments, deployed Ollama with Llama/DeepSeek for offline enterprise LLM. Developed AI agents converting natural language into Revit/AutoCAD actions.

RayOllamaLLMDeepSeekBIMGPU

Sep 2024 — Present

Lead Data Scientist & AI Engineer | Retail Analytics

DataPi (Asten collaboration)

Djerba, Tunisia

Built IRIS, a multi-agent Telegram AI assistant with Whisper V2, LLM reporting, anomaly detection, and forecasting. Developed pricing intelligence engines, retail dashboards, YOLOv8 prototypes for automated checkout, and IRIScraper for competitor pricing.

Multi-Agent AIWhisperYOLOv8MetabaseDocker SwarmPricing AI

Sep 2023 — Sep 2024

Senior Data Scientist & Software Developer

DataPi

Djerba, Tunisia

Led a full Azure to on-prem migration, rebuilding ingestion, ETL, reporting, MySQL warehousing, and Docker pipelines in Python. Built retail analytics dashboards and evaluated Kafka/Druid/Superset stacks for streaming insights.

Azure MigrationETLKafkaDruidPython

Sep 2022 — Sep 2023

Data Scientist & Software Developer

DataPi

Djerba, Tunisia

Built Azure ingestion/processing pipelines using Data Lake, Databricks. Automated fraud detection alerts via WhatsApp/email with Twilio. Implemented Azure Logic Apps for monitoring and alerting orchestration.

AzureDatabricksTwilioFraud Detection

Apr 2021 — Jun 2022

Research Internship — Brain-Computer Interfaces

Université de Moncton

Moncton, NB, Canada

Developed CNN classification algorithms for motor-imagery EEG signals. Published research paper on novel deep learning approaches for Brain-Machine Interfaces. Multi-class classification of motor imagination.

EEGCNNBCIDeep LearningResearch

Jan 2021 — Jul 2021

End of Studies Internship — NLP

Military Hospital of Instruction, Tunis

Tunis, Tunisia

Developed a support system for writing scientific articles for doctors. Fine-tuned GPT-2 model for biomedical text generation.

GPT-2NLPFine-tuningBiomedical

Jun 2020 — Aug 2020

Research Internship — COVID-19 NLP

Université de Moncton

Moncton, NB, Canada

Automatic deep learning recovery of COVID-19 related data for clinical decision support. Adjusted BERT model on the biomedical Cov-Dat-20 database.

BERTCOVID-19NLPClinical AI

Dec 2019 — Jul 2020

Associate Teacher — Deep Learning

ESPRIT Engineering School

Tunis, Tunisia

Delivered hands-on Deep Learning sessions using NVIDIA DLI labs (CNNs, RNNs, transfer learning, GPU workflows). Coached student teams on end-to-end Data Science projects.

NVIDIA DLITeachingCNNRNN

Tech Stack

Skills & Technologies

Bridging the gap between research and production.

🤖 AI / Machine Learning

LLMs (Llama, DeepSeek, GPT)95%

Computer Vision (YOLO, CNN)90%

NLP (BERT, GPT-2, Whisper)90%

Multi-Agent Systems88%

Time Series / Forecasting85%

💻 Data Engineering

Python (Pandas, Scikit, PyTorch)95%

SQL / MySQL / PostgreSQL90%

Azure (Data Lake, Databricks)85%

ETL / Data Pipelines90%

Kafka / Druid / Superset75%

⚙️ DevOps & Infrastructure

Docker / Docker Swarm92%

Ray (Distributed GPU)88%

Ollama / Local LLM Serving90%

Linux / Tailscale / Networking85%

Metabase / BI Dashboards92%

🔨 Software Development

Java / JavaFX / Java EE85%

PHP / Symfony / CodeIgniter80%

Android Development70%

Web Scraping (Selenium, BS4)88%

REST APIs / Microservices88%

Portfolio

Key Projects

Production systems built from research to deployment.

🤖

IRIS — Multi-Agent Retail AI

Telegram-based AI assistant with Whisper V2 voice input, LLM-powered PDF reporting, anomaly detection, top-10 forecasting, promo analysis, and diet-specific product recommendations.

Multi-AgentWhisperLLMTelegram

👁

YOLOv8 Automated Checkout

Computer vision prototypes for retail: product detection, counting, and brand classification at checkout. Trained on custom retail datasets.

YOLOv8Computer VisionRetail

🚀

BIM AI Agent

AI agents converting natural language prompts into Revit/AutoCAD actions. On-premise Ray GPU cluster with Ollama + DeepSeek for privacy-preserving inference.

RayOllamaBIMLLM Agent

📈

Pricing Intelligence Engine

Elasticity-based optimal pricing, real-time competitor tracking with IRIScraper (Auchan & Carrefour), and automated margin protection system.

ML PricingScrapingAnalytics

🧠

Brain-Computer Interface (BCI)

Novel CNN classification of motor-imagery EEG signals. Published research paper on deep learning approaches for brain-machine interfaces.

EEGCNNPublished Paper

Read the paper on MDPI ↗

🏥

Azure Cloud Analytics Platform

Full Azure pipeline: Data Lake, Databricks, Logic Apps, Twilio integration for automated fraud detection alerts and performance monitoring.

AzureDatabricksTwilio

Turning Data into Decisions

AI & ML Engineering

Data Analytics

Cloud & Infrastructure

Research & Teaching

Professional Experience

Skills & Technologies

🤖 AI / Machine Learning

💻 Data Engineering

⚙️ DevOps & Infrastructure

🔨 Software Development

Key Projects

IRIS — Multi-Agent Retail AI

YOLOv8 Automated Checkout

BIM AI Agent

Pricing Intelligence Engine

Brain-Computer Interface (BCI)

Azure Cloud Analytics Platform

Academic Background

Research Fellowship

Université de Moncton, Canada

Master's in Research

ISIMM — Université de Monastir

Engineering Degree

ESPRIT

Bachelor's Degree

ISIMM — Université de Monastir

Certifications & Publication

Fundamentals of Deep Learning for Computer Vision

Fundamentals of Accelerated Data Science with RAPIDS

What is Data Science?

A Novel CNN Classification Approach of Motor-Imagery EEG Recording

Get in Touch

Email

Location

LinkedIn