Responsibilities include:

• Led a monocular 4D reconstruction and representation learning project using 3D Gaussian Splatting, achieving SOTA depth estimation and comparable RGB reconstruction and tracking. Trained on (16-32)x A100 and (8-16)x H100 GPUs.
• Extended internal open-source system incorporating 3D visualization methods.

Responsibilities include:

• Led a navigation project for multi-legged robots utilizing inverse reinforcement learning and vision-based dynamics deep learning models.

Responsibilities include:

• Developed and led an open-source project to help diagnose diseases using AI, leading to a demo.
• Conducted research applying deep learning in the medical field on portable devices, leading to a published paper.
• Invented a new loss that preserves neural network equivariance under extreme transformations, which was subsequently patented (US Patent App. 17/170,745).

Responsibilities include:

• Led the development of text-based analytics solutions and dashboards for customer textual feedback and comments. This project reduced manual work by 43%, and enabled clients to make quick data-driven decisions.
• Managed strategy, product development, and client partnerships, including landing Zain, a major multinational telecommunications company, as an ongoing client.
• Currently supporting the development of the next solution with cutting-edge LLMs and RAG technologies.

Responsibilities include:

• Developed predictive machine learning models that analyse complex data and produce actionable insights to enhance customer understanding and user-centric product development.
• Conducted social network analysis (SNA) on calling data to identify communities of customers who interact frequently or share characteristics, and examine how these communities relate to ARPU levels and churn rates.
• Identify market trends to inform strategic marketing decisions and support data-driven strategies by providing comprehensive analytical insights.