🔬 PhD Researcher · UCBL/LIRIS ⚙️ Applied AI Engineer 💡 AI Consultant Open to full-time · Lyon / Paris

Ichrak Ennaceur

AI Engineer · AI Consultant & Applied Researcher in Generative AI

I design and build AI systems end-to-end — RAG pipelines, LLM evaluation, and recommendation models. PhD candidate at UCBL/LIRIS with 4 peer-reviewed publications and 6+ years delivering production-grade AI.

✉ Get in touch Download CV View work ↓

What I bring

4

Peer-reviewed Publications

Published at WISE 2025 and AICCSA 2025 on RAG-based recommendation systems, LLMs, and educational AI.

PhD

Industrial Research @ UCBL/LIRIS

Thesis on generative AI recommendation systems — bridging academic research and real-world deployment at Inokufu.

E2E

Full-Stack AI Engineering

I own the full pipeline — data ingestion, embeddings, retrieval, LLM generation, evaluation, and production deployment.

6+

Years Shipping Production AI

From call center NLP to RAG systems and AI agents — delivering tested, documented, production-ready code across industries.

6+
Years in AI/Data
4
Peer-reviewed papers
10+
Projects delivered
15+
AI skills mastered

Core Expertise

🤖 Generative AI & LLMs

Claude API OpenAI API Mistral Llama 3 Gemma Qwen DeepSeek RAG Prompt Engineering Fine-tuning (LoRA / QLoRA) Hugging Face In-Context Learning Transformers Structured Outputs

🤝 AI Agents & Orchestration

LangGraph LangChain n8n CrewAI AutoGen Multi-Agent Systems Tool Use / Function Calling MCP Agentic Workflows

📈 Evaluation, MLOps & Observability

LLM-as-a-Judge RAGAS DeepEval Evidently AI MLflow Weights & Biases Langfuse Arize Phoenix Hallucination Detection LLMOps A/B Testing Data Drift Detection

📊 ML & Recommendation Systems

BERT4Rec RecBole Sequential Models Collaborative Filtering Matrix Factorization Deep Learning PyTorch TensorFlow / Keras scikit-learn XGBoost / LightGBM Feature Engineering Model Evaluation

🔍 Vector Search & Retrieval

ChromaDB FAISS Pinecone Milvus Qdrant Weaviate pgvector Elasticsearch Hybrid Search Reranking Semantic Search

🗣️ NLP & Data Processing

spaCy NLTK Named Entity Recognition Text Classification Tokenization pandas NumPy Apache Spark dbt Apache Airflow

🛠️ Backend, Infra & Deployment

FastAPI Python Docker Kubernetes REST APIs PostgreSQL MongoDB Redis Kafka GitHub Actions CI/CD Terraform AWS / GCP Netlify / Vercel Git

🔐 AI Governance & Security

Responsible AI Bias Detection AI Safety Red Teaming Data Privacy (GDPR) Model Cards Explainability (XAI) Watermarking

Experience

AI Engineer & Research Scientist
Inokufu · Lyon Full-time Current 2022 – Present
Designing production RAG systems + recommendation engines + hallucination evaluation — while co-authoring research at UCBL/LIRIS
RAG BERT4Rec LLM-as-a-Judge ChromaDB FastAPI LangGraph PyTorch
Details ▾
  • Architected a hybrid BERT4Rec + RAG recommendation engine for educational content, combining vector search (ChromaDB) and LLM generation — deployed end-to-end in production via FastAPI
  • Built an LLM-as-a-Judge evaluation pipeline to detect and mitigate hallucinations in RAG outputs, improving system reliability and user trust
  • Designed scalable data pipelines for ingestion, embedding, and validation of large-scale educational content datasets
  • Contributed to the open-source Learning Records Converter (SCORM, IMS Caliper, xAPI), improving learning data interoperability across EdTech platforms
  • Co-authored 2 peer-reviewed papers (WISE 2025, AICCSA 2025) on generative AI recommendation systems for education
Lecturer — Data Science & AI
Université Claude Bernard Lyon 1 / Polytech Lyon Teaching Current 2024 – Present
Teaching data science, machine learning, deep learning, and business intelligence — with applied projects and LLM/RAG introductions
Machine Learning Deep Learning RAG Data Science Business Intelligence
Details ▾
  • Teaching data analysis, machine learning, deep learning, and business intelligence to engineering students at Polytech Lyon / UCBL
  • Designed and supervised end-to-end data science projects — guiding students through data preparation, modelling, and evaluation
  • Introduced students to LLM architectures and Retrieval-Augmented Generation (RAG) through applied hands-on projects
  • Mentored student teams on real-world AI/data challenges, from problem framing to production-ready solutions
Lead AI Engineer — Sales Forecasting Platform
MathildeCreations Freelance 6 months
Architect and developer of an AI-powered forecasting platform with an LLM explainability agent
FastAPI Hexagonal Architecture Docker LLM Agent CI/CD
Details ▾
  • Designed and delivered a full Python/FastAPI backend with hexagonal (clean) architecture — enabling high testability, modularity, and long-term maintainability
  • Built ML-powered forecasting endpoints for sales prediction and product potential scoring, integrated directly into client workflows
  • Implemented an LLM agent that explains predictions in natural language, making AI outputs interpretable for non-technical business stakeholders
  • Set up a Docker-first CI/CD pipeline with automated testing and linting, ensuring reproducible and reliable deployments
Generative AI Engineer — LLM Content Evaluation
ContentCrea Freelance 6 months
Automated multi-dimensional editorial quality scoring with LLM-as-a-Judge
LLM-as-a-Judge Claude API Prompt Engineering Evaluation
Details ▾
  • Designed a multi-dimensional LLM-as-a-Judge system using Claude API to automatically score editorial content on clarity, accuracy, tone, and SEO alignment
  • Built a generate → critique → refine pipeline that iteratively improves content quality with minimal human intervention, reducing manual review time
  • Extracted prompt optimization signals from real user correction patterns, continuously improving judge precision and reducing false positives
Lead Data Scientist — NLP & Audio AI
1waycom · Tunis Full-time 2020 – 2022
Led AI development for call center automation: Speech-to-Text, sentiment analysis, and real-time dashboards
Deep Learning NLP Speech-to-Text Sentiment Analysis Python
Details ▾
  • Led the design and delivery of an end-to-end auto-evaluation system for call center agents, covering productivity metrics, customer sentiment, and quality scoring
  • Trained and deployed Deep Learning models for Speech-to-Text transcription and audio-based sentiment analysis in a live production environment
  • Shipped Python microservices integrated with a Laravel frontend, serving real-time agent performance monitoring to operations teams
  • Built interactive management dashboards enabling live KPI tracking and data-driven decision-making across agent teams

Education

Université Claude Bernard Lyon 1 — LIRIS In progress
PhD in Artificial Intelligence · Industrial Thesis (CIFRE)
"Generative AI-based Recommendation Systems for Educational Platforms" — Research conducted in partnership with Inokufu, focused on RAG architectures, LLM evaluation, and sequential recommendation models.
RAG LLMs Recommendation Systems EdTech LIRIS Lab
2023 – 2026
📍 Lyon, France
University of Sfax
Master's in Data Science & Engineering
Machine Learning Data Engineering Statistics Deep Learning
2019 – 2021
📍 Sfax, Tunisia
University of Sfax
Bachelor's in Computer Science
Algorithms Software Engineering Databases Networks
2015 – 2019
📍 Sfax, Tunisia

Publications

CLARE: A Category-Aware RAG-Based Framework for Recommending Learning Objects
WISE 2025 Ichrak Ennaceur et al.
RAG Recommender Systems BERT4Rec EdTech
Prompt-Based Recommendation with In-Context Learning and Category-Enriched Modeling
🏆 Best Paper
AICCSA 2025 Ichrak Ennaceur et al.
LLM In-Context Learning Prompt Engineering NLP
CLARE : un cadre de recommandation d'objets d'apprentissage intégrant un RAG centré sur les compétences pour l'éducation et la formation professionnelle
EGC 2026 Ichrak Ennaceur et al.
RAG Skill-based Retrieval Learning Objects EdTech
Diagnosing Hallucinations in RAG-Based Educational Recommendation Systems: A Controlled Comparative Study
ISMIS 2026 Ichrak Ennaceur et al.
Hallucination RAG Evaluation Groundedness Evaluation Protocols

Academic Service

🏛️
Technical Committee Chair
AICCSA 2025
ACS/IEEE International Conference · 2025
🏛️
Technical Committee Chair
AICCSA 2026
ACS/IEEE International Conference · 2026
🏛️
Technical Committee Chair
APWebWAIM 2026
Asia Pacific Web & Web-Age Information Management · 2026

Projects & Portfolio

🎓
Course RAG API
Semantic course retrieval + LLM-powered recommendations for EdTech. Hexagonal architecture with pluggable FAISS/Qdrant/Milvus backends and OpenAI/Anthropic adapters.
FastAPI FAISS Sentence-Transformers OpenAI Anthropic
View Details → ⌥ Code
🏢
Enterprise Knowledge Assistant
Production-oriented RAG system for enterprise internal knowledge bases. Clean layered architecture, source-backed answers, Milvus vector retrieval. In progress.
FastAPI Milvus RAG Python
View Details → ⌥ Code
🔍
RAG Hallucination Evaluator
End-to-end framework for evaluating and correcting hallucinations in RAG pipelines. Uses Claude API as judge with an iterative correction loop.
Claude API LangGraph ChromaDB FastAPI
View Details → ⌥ Code
🎯
BERT4Rec + RAG Recommender
Hybrid recommendation system combining sequential models with retrieval-augmented generation for educational content.
BERT4Rec PyTorch RAG FAISS
View Details → ⌥ Code
📝
Career Signal
AI-powered job search operating system — content generation, LinkedIn scheduling & autopublish, networking outreach, job tracking, and technical watch. All in one agentic platform.
LangGraph Claude API OpenAI FastAPI Playwright Docker
View Details → ⌥ Code
🌐
GitHub Profile
GitHub profile with pinned projects, active contributions, and thorough documentation. Reflects expertise in generative AI.
Open Source Portfolio
Visit Profile →
📚
Modern Web Portfolio
This portfolio — built from scratch with HTML/CSS/JS. Interactive timeline, dark mode, animated hero, and deployed on Netlify.
HTML/CSS JavaScript Responsive Netlify
You're here →

Writing & Articles

View all on Medium →
🧠
Medium
AI Engineering · LLM Efficiency
Why Prompt Engineering Is a Crucial Step in Maximizing AI Efficiency
A practical deep-dive into prompt engineering as a core production skill — not a nice-to-have but a critical lever for AI system performance.
Ichrak Ennaceur Read on Medium →
✍️
AI Engineering · Prompt Design
Why Prompt Engineering Is the Skill No One Told You to Master
It's not about talking to an AI. It's about thinking precisely — and that changes everything about how you build with LLMs.
Ichrak Ennaceur Read →
📚
Medium
RAG · LLM Architecture
Building Smarter LLMs with RAG: From Documents to Answers
How Retrieval-Augmented Generation grounds LLM responses in real documents — architecture, retrieval pipeline, and production trade-offs.
Ichrak Ennaceur Read on Medium →
📊
Medium
AI Evaluation · KPIs · Opinion
The Honest AI KPI Nobody Puts on the Slide
We're measuring the wrong things — and we know it. A critical look at the vanity metrics dominating AI project reporting and what actually matters.
Ichrak Ennaceur Read on Medium →

Beyond Technical Skills

Structured around the DocPro 24 competencies framework — the professional profile of PhD doctors.

🎯
Core Expertise
  • Expertise & Methods
    Specialized in Generative AI, RAG, LLMs and recommendation systems — with 4 peer-reviewed publications and end-to-end production deployments.
  • Evaluation
    Designed LLM-as-a-Judge pipelines, groundedness evaluation protocols, and hallucination analysis frameworks for RAG systems.
  • Knowledge Sharing
    Lecturer at Polytech Lyon / UCBL: supervising end-to-end data science projects, teaching ML, deep learning and RAG architectures.
🤝
Personal & Relational Qualities
  • Analysis & Critical Thinking
    Translates complex problems (RAG hallucinations, sequential recommendation) into system architectures and peer-reviewed research contributions.
  • Communication
    Scientific publications, Medium articles, teaching, and presenting to both technical and non-technical audiences. Trilingual: French, English, Arabic.
  • Commitment & Integrity
    Simultaneously running an industrial PhD, a full-time engineering role, and freelance missions — with documented deliverables and clean code at every step.
📊
Activity Management & Value Creation
  • Project Management
    Led AI projects end-to-end — from design to production (Inokufu, MathildeCreations, ContentCrea) — working across research and product teams.
  • Decision Making
    Architecture choices (hexagonal, hybrid RAG, multi-agent), tech stack selection, and research/engineering trade-offs under real-world constraints.
  • Delivering Results
    4 publications, multiple AI systems in production, 2 freelance projects delivered — each with CI/CD, tests, and full documentation.
🚀
Strategy & Leadership
  • Strategy
    Designing AI roadmaps — from business problem identification to system architecture, including evaluation metric selection and stakeholder alignment.
  • Leadership
    Lead AI Engineer on freelance projects, coordinator between research and product teams, mentor for data science and AI students.
  • Change Management
    Supporting the adoption of Generative AI in industrial settings — explaining limitations, risks, and integration best practices to technical and business stakeholders.

Outside of Work

What keeps me balanced and curious beyond the screen.

🎾
Tennis
Strategy on the court, precision under pressure.
🏊
Swimming
Clarity of mind through rhythm and flow.
⛺
Camping
Disconnecting to reconnect with what matters.
🎨
Painting
Creativity without constraints — a different kind of problem-solving.

Let's Work Together

Open to full-time positions in AI Engineering or Applied Research — Lyon, Paris, or remote. Let's talk about what you're building.

✉ Send me an email 💼 LinkedIn ⌥ GitHub 📄 Download CV
📍 Lyon, France 🎓 PhD Candidate · UCBL / LIRIS ✅ Open to full-time · Lyon / Paris / Remote