I am an AI and Machine Learning Engineer specializing in agentic systems, retrieval-augmented generation (RAG), natural language processing, and computer vision. I build intelligent applications that combine robust model evaluation, backend engineering, and practical product integration.
I have architected LangGraph-based orchestration and RAG pipelines for a medical assistant, including safety-focused routing, semantic retrieval, real-time web search, and LLM-driven decision paths. My work emphasizes reliable tool use, traceable workflows, and contextually relevant outputs.
I have hands-on experience developing and evaluating Transformer-based NLP systems for question answering, summarization, semantic search, and named entity recognition. I also work with Python machine learning ecosystems including PyTorch, TensorFlow, and scikit-learn.
In computer vision, I led the evaluation and integration of multilingual lip-reading models. I verified 99.1% word-level accuracy for an English model aligned with Meta's published benchmark and contributed to research evaluation for Arabic and sentence-level lip-reading systems.
I also have backend AI engineering experience with FastAPI, Node.js/Express, Docker, PostgreSQL, authentication, API usage metering, billing workflows, Stripe webhooks, and background processing. I am fluent in English and a native Arabic speaker.
I conduct agent error analysis and function-calling validation for multi-step reasoning and API logic. I audit agent behavior to identify reliability issues and improve execution quality.
I evaluate model alignment by comparing recursive agent outputs against corrected versions. This work supports more consistent decision-making and improved model reliability.
I completed parallel AI Fluency and Backend AI Engineering internship tracks, delivering more than 12 projects using FastAPI, Node.js/Express, and Docker. I built supporting AI and backend solutions including RuleGuard AI, Decision Flow Studio, JWT authentication, report processing, PDF generation, and web scraping.
For my capstone, I architected a multi-tenant usage metering and billing engine with idempotent request handling, tenant-level monthly quotas, and integer micro-unit cost calculations. I synchronized subscription state through verified Stripe test-mode webhooks and implemented a background reconciliation process with retries and terminal-failure handling.
I developed Transformer-based BERT and T5 solutions for question answering and text summarization. I optimized model performance using ROUGE and F1 evaluation metrics.
I also built semantic search and named entity recognition pipelines using Sentence Transformers and spaCy. These pipelines supported automated ranking and entity extraction workflows.
I applied transfer learning techniques to build neural networks for NLP and computer vision tasks. I worked across machine learning workflows involving model development and experimentation.
I streamlined data analysis and model evaluation pipelines to support more efficient experimentation and performance assessment.
I developed five machine learning solutions spanning supervised learning, unsupervised learning, and reinforcement learning. I implemented end-to-end workflows using Python, scikit-learn, and TensorFlow.
My work covered practical model development, experimentation, and evaluation across multiple machine learning problem domains.
Bachelor's Degree, Artificial Intelligence
Major in Artificial Intelligence. GPA: 3.72.
This professional hasn’t added portfolio projects yet.
This professional hasn’t listed any services yet.
