Uzma Aftab

AI/Machine Learning Engineer with a BS in Artificial Intelligence and hands-on experience building multimodal deep learning systems spanning computer vision, audio signal processing, NLP, and biosignal (EEG) analysis. Designed and deployed VERITAS AI, a four-modality lie-detection system combining ResNet-18, BiLSTM, and EEGNet architectures with weighted fusion, achieving up to 86.69% single-modality accuracy. Skilled in Python, TensorFlow, and PyTorch, with proven ability to take AI models from research and preprocessing through to FastAPI-based production deployment.

Seeking to bring applied AI engineering expertise to a UAE-based technology or innovation-driven team.

Artificial Intelligence Intern - Smart Tech Solutions - Islamabad, Pakistan

(2025-06 - 2025-08)

• Developed a real-time AI chatbot system applying Natural Language Processing techniques to handle conversational query flows.
• Applied Machine Learning, Deep Learning, Computer Vision, and NLP methods across multiple academic and practical projects during an 8-week internship.
• Executed end-to-end data preprocessing, feature extraction, model training, and evaluation to support functional AI prototypes.
• Collaborated with a cross-functional engineering team to integrate trained AI models into working systems.
• Authored technical documentation and evaluation reports to support internal knowledge-sharing and project review.

Bachelor of Science - Artificial Intelligence - University of Kotli AJK (2021-12 - 2025-12)