Senior AI Engineer · Riyadh, Saudi Arabia
5+ years of AI/ML and software delivery across LLM-powered platforms, RAG pipelines, computer-vision models, and cloud-ready AI services.
About
I combine applied AI research, full-stack engineering, and deployment experience to deliver reliable AI products end to end. My work covers data preparation, PEFT/LoRA fine-tuning, RLHF-style evaluation, prompt evaluation, quantization, ONNX export, RAG evaluation, secure APIs, and production rollout.
I am strongest where AI models meet real products: Python ML services connected to ASP.NET Core and React, agent workflows coordinated with n8n and Elsa Workflows, vector memory, cloud deployment, CI/CD, observability, and human-in-the-loop operations.
Experience
Selected Projects
Enterprise support automation platform for high-volume e-commerce, using intent classification, RAG, multilingual chat, agent takeover, live operations, Firestore monitoring, and BigQuery analytics.
Research intelligence system for paper discovery, deep analysis, Scientific Code Forge generation, repository mapping, claim auditing, and multi-agent debate reports.
View repositoryMultilingual, dialect-aware assistant combining Whisper v3, Arabic dialect classification, semantic matching, GPT-family reasoning, WebSockets, and orchestration flows.
Retrieval-grounded NL-to-SQL assistant with schema-aware guardrails, safe execution constraints, and validation checks that reduce hallucinated tables and columns.
CNN-based Arabic character recognition trained on 120,000 samples, achieving 85% top-1 accuracy, 59% model-size reduction, and 2.3x CPU inference speedup.
Hybrid image-captioning system that routes between fast template captions and richer LLM captions using confidence, entropy, color, and complexity signals.
View repositoryMultitask OCR and verification pipeline for mixed Arabic/Latin prescriptions, seal detection, dosage standardization, 35% fewer interpretation errors, and 92% character-level accuracy.
Recruitment tool using TF-IDF, NLTK, KNN/SVM/Random Forest models, role-based access, and automated candidate ranking that reduced shortlisting time by 40%.
View repositoryTechnical Expertise
Python, C#, JavaScript/TypeScript, SQL, MATLAB
PyTorch, TensorFlow, Keras, scikit-learn, Pandas, NumPy, Hugging Face Transformers, ONNX
Embeddings, chunking, retrieval, reranking, RAGAS, structured outputs, prompt evaluation, guardrails
ChromaDB, FAISS, Pinecone, Weaviate, Vertex AI, OpenAI, Anthropic
LangChain, n8n, Elsa Workflows, tool-calling, planner/executor patterns, long-context memory
FastAPI, Flask, ASP.NET Core, Entity Framework, REST APIs, WebSockets, Uvicorn, Gunicorn
Docker, Git, CI/CD, IIS, AWS, GCP, PythonAnywhere, Heroku, MLflow, W&B, Airflow, Prefect
OpenTelemetry, Prometheus, Grafana, PyTest, OAuth2, JWT, AWS Secrets Manager, GCP Secret Manager
Education
University of Hull · Hull, United Kingdom
Focus on computer vision, deep learning, and LLM applications.
University of Kalamoon · GPA 3.08 / 4.00 (83 / 100)
Graduated with distinction and achieved the second-highest rank in class.
Yousef Seman Industrial High School · GPA 92 / 100
Top academic performer and ranked fifth in the governorate in final 12th-grade exams.
Credentials
Contact
Reach out for LLM systems, RAG reliability, AI agents, computer vision, MLOps, or full-stack AI delivery.