Mr. Saad Azhar SCIT - BNU

Mr. Saad Azhar is a Senior Lecturer at the School of Computer & Information Technology (SCIT), Beaconhouse National University (BNU), Lahore. He has over a decade of professional experience spanning machine learning, deep learning, data engineering, and full-stack software development.

Currently, he is pursuing a PhD in Machine/Deep Learning and serves as a Machine Learning Scientist for Comparables AI, a Finnish startup. He has also worked as a data engineer in an Australian e-commerce startup and has previously served as a lecturer at University of Central Punjab. His experience covers theoretical concepts and practical implementation for AI, web development, and software engineering projects.

• PhD (Course Work Completed) in Computer Science, National University of Computer and Emerging Sciences (FAST), Lahore, Pakistan (2021 – Present)
• MS in Computer Science, National University of Computer and Emerging Sciences (FAST), Lahore, Pakistan

• Senior Lecturer, School of Computer & Information Technology, Beaconhouse National University (BNU), Lahore
• Machine Learning Scientist, Comparables AI / ByteSol, Helsinki, Finland (Remote Pakistan) (2021 – Present)
• Software/Data Engineer, Prowisetech / Mega-e-Store, Sydney, Australia (Remote Pakistan) (2020 – 2022)
• Lecturer, University of Central Punjab, Lahore, Pakistan (2017 – 2021)
• Full-Stack Developer / Software Engineer, Intuitive App Studio and Makeityourweb / 3S Solutions Pvt Ltd – Various Roles

• Generative AI, Data Extraction, Summarization, RAG with LLMs (ChatGPT, Gemini, Vertex AI)
• Fine-tuning NLP models, embeddings with Llama 2, Sentence Transformer (MPNET), BERT
• Python (Numpy, Pandas, Scrapy, Matplotlib, Seaborn, Sklearn), FASTAPI, PyTorch
• Machine Learning: PCA, FA, Naive Bayes, SVM, Decision Tree, Random Forest, MLP, K-means, KNN
• Data Sources: ElasticSearch, PostgreSQL, MySQL | Cloud Platforms: AWS, Azure
• Web Development: HTML, CSS, JavaScript, PHP, Laravel, WordPress, Magento
• Deployment: Docker, FASTAPI, GPU-based embedding generation, inference endpoints

• Comparables AI: Data extraction, summarization, knowledge graph generation, user intent detection, BERT pretraining and fine-tuning of embeddings for search & recommendation systems.
  https://www.comparables.ai/
• UAE General Women’s Union: AR-based application for royal family awards with web admin panel, Android & iOS apps.
  Android App | Web App
• Bizbench: Accounting software for business benchmarking and reporting using PHP, WordPress, ExtJS.
  http://bizbench.com
• Utah Hogle Zoo: Website redevelopment and admin panel implementation for zoo management using PHP, WordPress, ExtJS.
  http://www.hoglezoo.org/

• Email Address: saad.azhar@bnu.edu.pk