GeezerVison

🤖 AI/ML Engineer · Building production AI systems that ship to real users

AI Accuracy
_{Production Invoice System}

Users Impacted
_{FinTech Banking Platform}

Company Interest
_{Healthcare AI Inquiry}

Collaborations
_{Global Requests}

`whoami`

#!/usr/bin/env python3
"""
Ahlem Laajili - AI/ML Engineer
Building production AI that delivers measurable business impact.
"""

class AIEngineer:
    def __init__(self):
        self.name = "Ahlem Laajili"
        self.title = "AI/ML Engineer | Python Developer | Full Stack Engineer"
        self.location = "Tunisia 🇹🇳 → Remote Worldwide 🌍"
        
        self.education = {
            "PhD": {"field": "Artificial Intelligence", "school": "SUP'COM", "status": "Admitted"},
            "MSc": {"field": "Computer Science Engineering", "school": "ESPRIT", "rank": "Graduated with High Honors"},
            "BSc": {"field": "Computer Science & Networking", "school": "ISTIC", "rank": "Top 3 🏆"},
        }
        
        self.current_role = {
            "company": "BUSINESS & AI",
            "role": "AI Full Stack Engineer",
            "impact": "Engineering production-grade AI solutions combining computer vision, NLP, and deep learning"
        }
        
        self.achievements = [
            "🎯 95.4% accuracy in production AI systems",
            "🌍 50+ international collaboration requests",
            "🏆 ESPRIT Competition Finalist (9th Edition Bal des Projets)",
            "📺 1.5K+ YouTube tutorial views",
            "⭐ 400+ GitHub stars across projects",
        ]
        
        self.specializations = ["Document AI", "Computer Vision", "NLP", "LLM Fine-tuning", "MLOps"]
        
    def philosophy(self):
        return "I build AI that works in production, not just in notebooks."

me = AIEngineer()
print(me.philosophy())  # Let's build something extraordinary together.

🌟 What Sets Me Apart

🤖 Production AI Systems

┌─────────────────────────────────────┐
│  Document AI    │ 95.4% Accuracy    │
│  Computer Vision│ Real-time @ 30fps │
│  NLP Pipelines  │ <100ms latency    │
│  LLM Fine-tuning│ LoRA/QLoRA/GGUF   │
│  MLOps          │ End-to-end deploy │
└─────────────────────────────────────┘

Document Intelligence → Automated data extraction and understanding pipelines
Deep Learning Models → Custom vision-based AI solutions
Natural Language Processing → Text understanding and information extraction systems
AI Applications → Production-ready intelligent automation workflows

⚡ Engineering at Scale

┌─────────────────────────────────────┐
│  FinTech APIs   │ 130K+ users       │
│  DB Migrations  │ 15+ production    │
│  Async Systems  │ Redis + Celery    │
│  CI/CD          │ Jenkins + Docker  │
│  Compliance     │ PCI-DSS ready     │
└─────────────────────────────────────┘

Django/FastAPI → High-throughput APIs
PostgreSQL → Complex migrations at scale
Redis → Caching & message queuing
Docker → Containerized deployments

🔬 Featured Projects

📄 Invoice Smart Learning

95.4% Accuracy | Production System

Continual learning system for document AI with hard example mining, active learning, and drift detection. 69.52% faster with CUDA optimization.

LayoutLMv3 PyTorch FastAPI Label Studio

🏥 AI Healthcare Assistant

ESPRIT Finalist | 1.1K+ Views | 100+ ⭐

Healthcare platform with face recognition, lung disease classification (X-ray AI), blockchain medical records, and intelligent chatbots.

TensorFlow Computer Vision Blockchain React

🎯 LLM Fine-tuning Toolkit

LoRA → QLoRA → GGUF Pipeline

Complete fine-tuning pipeline from dataset prep to Ollama deployment. Supports Llama 3, Mistral, Gemma, Qwen, and any HuggingFace model.

PyTorch HuggingFace PEFT bitsandbytes

🔍 RAG Production Template

Enterprise-Ready | Hybrid Search

Beyond tutorials: semantic chunking, hybrid search (Vector + BM25 + Reranking), RAGAS evaluation, Redis caching, and Kubernetes configs.

LangChain FastAPI Qdrant Redis

📦 Open Source Portfolio

🤖 AI/ML Projects

Project	Description	Tech Stack
Invoice Smart Learning	95.4% accuracy document extraction with continual learning	`LayoutLMv3` `CUDA` `FastAPI`
AI Healthcare Assistant	Healthcare AI with face recognition & blockchain	`TensorFlow` `React` `Flask`
LLM Fine-tuning Toolkit	LoRA/QLoRA training → GGUF/Ollama export pipeline	`PyTorch` `PEFT` `HuggingFace`
RAG Production Template	Enterprise RAG with hybrid search & RAGAS evaluation	`LangChain` `FastAPI` `Qdrant`
AI Agents Starter	Production agent templates with tool calling & memory	`LangChain` `OpenAI` `Tools`
ML Concepts Explained	Interactive ML tutorials: entropy → transformers	`PyTorch` `NumPy` `Jupyter`

👁️ Computer Vision Projects

Project	Description	Tech Stack
YOLO Real-time Detection	80+ object classes with Gradio UI, webcam & video	`YOLOv8` `OpenCV` `Gradio`
Face Emotion Detector	7 emotions with 95% accuracy, real-time	`DeepFace` `OpenCV` `Gradio`
KYC Liveness Detection	Anti-spoofing with face challenges: blink, turn, smile	`MediaPipe` `OpenCV` `Flask`
Hand Gesture Recognition	Real-time hand tracking for HCI applications	`MediaPipe` `TensorFlow`
License Plate Reader	ANPR system with OCR extraction	`OpenCV` `Tesseract` `EasyOCR`
AI Background Remover	Automatic background removal for images	`U2-Net` `OpenCV` `Gradio`

🛠️ Templates & Boilerplates

Project	Description	Tech Stack
FastAPI ML Template	Production-ready ML API with Docker & tests	`FastAPI` `Docker` `Pytest`
Django Production Boilerplate	Scalable Django setup with best practices	`Django` `PostgreSQL` `Redis`
Document AI Pipeline	Modular document processing framework	`LayoutLM` `Tesseract` `FastAPI`
RL CartPole Agent	DQN with experience replay, complete RL tutorial	`PyTorch` `Gymnasium`

🎬 Technical Content

📺 YouTube — Production AI Tutorials

Join 1K+ engineers learning from actual implementations.

Healthcare AI Platform
Face recognition + Lung AI + Blockchain
🎯 1.1K+ views ⏱️ 3:00

95.4% Invoice Extraction
LayoutLMv3 in production
🎯 283 views ⏱️ 4:53

Technical Tutorials
Django, ML, DevOps
🎯 Growing ⏱️ Varies

🛠️ Tech Stack

🐍 Languages

🤖 AI / ML Frameworks

📊 Data & Visualization

⚙️ Backend & Data

🎨 Frontend

🚀 DevOps & Cloud

💬 What People Say

Real feedback from clients, founders, and industry leaders across 10+ countries.

"I came across your invoice processing demo and would be interested in discussing a potential project for a healthcare company with revenues of $40m each month..." — Mike Vezzoli, Healthcare Company Executive 🏥	"She writes outstanding documentation (probably the best I've seen out of 100+ candidates). If you are hiring - please reach out to her." — Lyra Solar Hiring Manager 🌍
"Ahlem demonstrated outstanding competencies in AI, using advanced technologies such as NLP and deep learning. I highly recommend her for any Data Science role." — Dr. Sonia Mesbah, ESPRIT 🎓	"I strongly recommend her to pursue doctoral work. Her multidisciplinary training, enthusiasm, and methodical character are tremendous assets." — Prof. Saber Dakhli, CentraleSupélec Paris 🇫🇷

🤝 Let's Collaborate

╭────────────────────────────────────────────────────────────────────────╮
│                                                                        │
│   🤖 AI Integration      →  Document AI, Computer Vision, NLP, LLMs   │
│   📺 Technical Content   →  Tutorials, workshops, documentation       │
│   🔍 Code Review         →  ML systems, Python architecture           │
│   💼 Consulting          →  Production ML system design               │
│                                                                        │
╰────────────────────────────────────────────────────────────────────────╯

💡 Building AI systems that work in production, not just in notebooks.

_{If you find my projects useful, consider ⭐ starring the repos!}