Mostafa Themar

Working as a freelance Machine Learning Engineer on various Arabic platforms including Khamsat, Mostaql, and others. Delivered multiple AI and data projects involving data extraction, automation workflows (n8n), voice data processing, and model-based analytics. Collaborate with clients to build customized AI solutions, improve business efficiency, and automate repetitive tasks.

Worked on automation workflows using n8n, including projects involving voice data recording and processing. Collaborated with banking systems to handle secure data operations and integrations. Also developed data extraction pipelines to support accounting and financial analysis.

Developed a RAG-based semantic search system for Arabic legal PDFs using AraBERT embeddings and Chroma vector store. demonstrating expertise in machine learning model design and fine-tuning

• Implemented an AI-driven business intelligence platform with CrewAl multi-agent workflows, reducing data hallucinations by 40% through automated extraction, validation, and analysis. • Developed robust web scrapers and integrated real-time pipelines using Groq's Llama3-70B, achieving 95% accuracy and sub-15-sec on insight delivery, reflecting collaborative efforts akin to ETL pipeline support.

Worked on data and image extraction from PDF documents to support automated data processing and analysis workflows. Built and optimized data pipelines for structured information retrieval and transformation. Collaborated with the team to improve data accuracy and efficiency using machine learning and automation tools.