Hi, I'm Jihad Akbar

Consulting role delivering data science and AI projects for five enterprise clients across different industries.

• Reviewed AI model predictions on large volumes of daily image and video data, sorting them by risk level to support safety compliance and model reliability.
• Labelled images with bounding boxes and segmentation and fed structured feedback back to the ML team, lifting model accuracy from 80% to 95%.

• Built a reporting pipeline from spreadsheets into a live dashboard, turning daily sales data into revenue and service recommendations for the sales team.
• Developed a booking cancellation prediction model reaching 86% AUC-ROC, helping the team act on high-risk reservations before they were lost.

• Developed a medical image classifier covering four eye conditions, with thorough image quality checks before training.
• Fine-tuned several deep learning models, selected the best at 83% accuracy, and deployed it to production as a web app.

• Annotated and reclassified large image datasets to support machine learning model development, keeping annotation accuracy and data quality high.
• Reviewed the work of around 30 annotators at 90-100% accuracy, introducing QA processes and feedback that raised consistency across the team.