I am AEK Meddad, a Machine Learning Engineer focused on turning ideas into dependable AI products. My work spans the full delivery path: data pipelines, model development, APIs, deployment, evaluation, and modern LLM applications.
The model is one component of the product. Useful AI combines sound data practices, machine learning, backend engineering, observability, security, and a clear user experience.
| I am building | I am strengthening |
|---|---|
| AI consent and approval infrastructure | Kubernetes and cloud infrastructure |
| Healthcare prior-authorization automation | MLOps, CI/CD, and model monitoring |
| AI evaluation and advanced RAG workflows | Agentic and distributed systems |
Explainable customer-churn prediction platform.
Stack: Python · XGBoost · Scikit-learn · SHAP · Flask |
AI consent infrastructure for high-impact actions.
Stack: FastAPI · LangGraph · PostgreSQL · Docker |
Healthcare Prior Authorization Intelligence System.
Stack: Python · FastAPI · LangGraph · ChromaDB · Docker |
AI study assistant for educational video content.
Stack: Python · FastAPI · LangChain · PostgreSQL · Railway |
|
Machine learning Model development · Explainability · Computer vision · Deep learning · Evaluation |
LLM systems RAG · LangChain · LangGraph · Multi-agent workflows · AI governance |
Product engineering FastAPI · Flask · PostgreSQL · Docker · CI/CD · Deployment |
- Build products, not notebooks. A model should solve a real user problem in a usable workflow.
- Make decisions observable. Measure performance, monitor systems, and explain predictions.
- Prefer deliberate simplicity. Start with the smallest reliable system, then iterate from evidence.
- Keep people in control. High-impact AI needs meaningful approval, safety, and governance.
- Ship, learn, improve. Deployment and feedback complete the engineering loop.
| Period | Focus | Milestones |
|---|---|---|
| 2023 | Programming foundations | Python, data structures, and algorithms |
| 2024 | Core machine learning | Classical ML, deep learning, computer vision, and NLP |
| 2025 | Production systems | FastAPI, Docker, LLM applications, and RAG systems |
| 2026 | AI infrastructure | Multi-agent systems, evaluation, Venta, and HPAIS |
| Area | Status | Progress |
|---|---|---|
| Production machine learning | Complete | ████████████ 100% |
| Deep learning | Complete | ████████████ 100% |
| Retrieval-augmented generation | Complete | ████████████ 100% |
| Multi-agent AI systems | Complete | ████████████ 100% |
| AI evaluation framework | In progress | ████████░░░░ 65% |
| MLOps and CI/CD | In progress | ██████░░░░░░ 50% |
| Kubernetes | In progress | ████░░░░░░░░ 35% |
| AWS cloud | In progress | ████░░░░░░░░ 30% |
| Open-source contributions | In progress | ██████░░░░░░ 45% |
| Project | Focus |
|---|---|
| AI Agent Evaluation | Hallucination, reasoning, toxicity, and response-quality evaluation |
| AI Job Search Agent | Autonomous job-search workflows using LangChain |
| Gymshark Recommendation System | Product recommendation engine |
| Hand Gesture Recognition | Real-time computer vision with OpenCV and MediaPipe |
| Face Mask Detection | TensorFlow CNN classifier |
| AI News Summarizer | Document summarization using NLP |
| Sudoku Solver | Backtracking search algorithm |
Contribution snake
- Machine Learning Engineering and AI Engineering roles
- Startup and remote opportunities
- Open-source collaboration and technical speaking
I am interested in building AI systems that people can trust, understand, and use. Feel free to reach out through email or LinkedIn.