Ilyes Ben Khalifa

About Me

A brief overview of who I am and what drives my work

I'm a Senior Data Scientist with deep expertise in applied AI, predictive modeling, and data-driven decision systems. Currently at Ominimo Insurance in the Netherlands, I lead pricing analysis and develop machine learning models for tariff optimization and risk segmentation.

My research spans software analytics, AI systems, and NLP — resulting in peer-reviewed publications in top-tier Q1 and Q2 journals including Information and Software Technology, IEEE Software, and Journal of Systems & Software.

I hold a Bachelor's in Information Technology from Tunis Business School. Alongside industry work, I maintain an active freelance presence on Upwork with a 100% Job Success Score and a Top-Rated profile, delivering AI solutions to clients worldwide.

5 Peer-Reviewed Publications

2× Q1 Journal Papers

3+ Years Industry Experience

100% Upwork Job Success Score

Research Publications

Peer-reviewed work published in Q1/Q2 journals and international venues, spanning software analytics, AI systems, and NLP

Q1 IF 3.862 Information & Software Technology

LARK: Python-Specialized License Analysis with RAG and Knowledge Graphs

I. Ben Khalifa, M. Ben Messaoud, M. W. Mkaouer

Information and Software Technology, Special Issue on Regulatory Compliance in Software Engineering · 2026

In Press — Link Coming Soon

Q2 IF 3.0 IEEE Software

LARK in Action: Smarter License Checking with AI, RAG, and Knowledge Graphs

I. Ben Khalifa, M. Ben Messaoud, M. W. Mkaouer

IEEE Software Magazine · 2026

View on IEEE Xplore

Q1 IF 4.1 Journal of Systems & Software

Hierarchical Multi-label Classification for Concrete Defects: An Industrial Case Study

M. Ben Messaoud, A. Nour, I. Ben Khalifa, M. Tounsi, M. W. Mkaouer

Journal of Systems & Software, 231:112588 · 2026

View on ScienceDirect

Q2 IF 2.6 IT Professional

Detecting Software Defects with Hierarchical Multilabel Classification: Insights from an Industrial Case Study

M. Ben Messaoud, A. Nour, I. Ben Khalifa, M. Tounsi, M. W. Mkaouer

IT Professional, 27(5):31–37 · 2025

View on IEEE Xplore

ERA Ranking B ICSOC 2024 · Springer LNCS

UEwMT: Leveraging User Experience and Deep Learning-Driven Methodology for Evaluating Machine Translation Services

K. Al Sharou, M. K. Jamei, I. Ben Khalifa, S. Missaoui, M. Ben Messaoud, J. Moorkens

Service-Oriented Computing – ICSOC 2024 Workshops, LNCS vol. 15834, Springer · 2024

View on Springer

Work Experience

Professional roles spanning insurance AI, LLM engineering, NLP research, and freelance consulting

Ominimo Insurance May 2025 – Present

Data Scientist I

Amsterdam, Netherlands · Remote

Lead pricing analysis for the Netherlands insurance portfolio, driving tariff optimization and competitive positioning.
Develop and deploy machine learning models for insurance coverages, risk scoring, and tariff calibration.
Run portfolio simulations to adjust pricing strategies and improve risk segment distribution.

UBIAI Jan 2024 – May 2025

Data Scientist & AI/NLP Content Writer

California, USA · Remote

Led fine-tuning of state-of-the-art LLMs (LLAVA, LLaMA 2/3, Mixtral, Qwen, DeepSeek Gemma) using RLHF, DPO, KTO; optimized deployment with LoRA, QLoRA, and OLLAMA.
Built LLM agents for NER and Relation Extraction that outperformed baseline zero-shot models, reducing document annotation time by 80% (from 1 min to 10 sec).
Architected multiagent workflows with CrewAI and LangGraph; scaled LLM infrastructure on AWS SageMaker, EC2, and GCP Cloud Run.
Designed ML training pipelines on Azure Databricks with PySpark; integrated SpaCy, BERT, and LayoutLM into Django-based annotation platforms.
Authored technical articles on Reinforced Fine-Tuning (ReFT) and comparative analyses of RLHF vs. RLAIF.

Freelance — Upwork 2023 – Present

AI/ML Developer · Top-Rated · 100% JSS

Remote

Delivered high-impact ML, neural network, and prompt engineering projects with consistent 5.0 client ratings.
Built generative AI solutions and RAG systems tailored to business needs, including LangGraph and CrewAI-based multiagent systems.
Designed robust ETL pipelines and integrated OpenAI APIs into scalable production applications.

Business & AI Nov 2022 – Jan 2024

Data Scientist

Ben Arous, Tunisia

Engineered NLP pipelines with LLMs for NER, achieving 95% accuracy extracting structured data from over 1 million web pages.
Built an end-to-end web scraping and ETL reporting system, boosting data extraction efficiency by 70%.
Led R&D on multimodal ML for speech-to-text and audio diarization, delivering a 40% accuracy improvement over 10,000+ hours of audio.
Architected a predictive analytics platform that improved forecast accuracy by 30% and reduced operational costs by 15%.

The Sparks Foundation Aug 2022 – Nov 2022

Data Science & Business Analytics Intern

Remote

Built scalable preprocessing pipelines with Pandas and NumPy, improving data quality by 40% and reducing analysis time by 25%.
Deployed classification and regression models with Scikit-Learn and XGBoost, achieving a 15% improvement over baselines.
Implemented NLP solutions processing 10,000+ documents and 500+ hours of audio, boosting information extraction efficiency by 30%.

PARADA Agency Jan 2022 – Aug 2022

Junior Machine Learning Engineer

Paris, France · Remote

Developed deep learning models for object detection and recognition using TensorFlow and PyTorch.
Implemented ML solutions for object tracking and pose estimation, enhancing accuracy and reducing latency.
Led data cleaning, augmentation, and labeling efforts for computer vision training datasets.

Skills & Technologies

A broad technical stack spanning ML, LLMs, cloud infrastructure, and data engineering

Languages & Data

Python SQL Git Pandas NumPy Matplotlib Seaborn

ML / DL Frameworks

TensorFlow PyTorch Scikit-learn XGBoost SpaCy BERT LayoutLM

LLMs & NLP

LangChain LangGraph CrewAI OLLAMA RAG LLM Fine-Tuning RLHF / RLAIF DPO / KTO LoRA / QLoRA NER REL Prompt Engineering

Cloud & Infrastructure

AWS SageMaker AWS EC2 GCP Cloud Run Azure Databricks PySpark Django

Data Engineering

ETL Pipelines Web Scraping Beautiful Soup Scrapy OpenAI API

Domain Expertise

Insurance Pricing Tariff Modeling Portfolio Simulation Algo Trading Research & Development Project Management