CV

Machine Learning Engineer

• Developed a multi-class ad policy violation classifier (50+ classes), increasing precision from 60% to 91% and reducing harmful ad exposure for 2B+ daily users.
• Built multiple AI detectors for malicious Chrome extensions, achieving over 95% precision at 50% recall.
• On my own initiative, improved LLaMA’s mathematical reasoning accuracy by 2%.
• On my own initiative, optimised Meta Quest’s eye-tracking algorithm achieving a 1% mAP improvement, enhancing gaze estimation and user experience across millions of devices.

Computer Vision Engineering Lead

• Established Petronas’ First CV Team & Scaled AI Adoption.
  • Built and led Petronas’ first Computer Vision team, growing it to 12 ML engineers.
  • Organised and led the internal Data Science Academy, training 100+ professionals in AI/ML.
  • Drove AI adoption across Petronas’ 10,000+ workforce through the Citizen Analytics program.
• Optimized Security Manpower with AI-Driven Video Analytics.
  • Led full deployment, including Linux-based server setup, optimizing manpower by ∼30%.
  • Built and deployed real-time surveillance automation across hundreds of cameras for Petronas Twin Towers & KLCC.
  • Developed CNN-based tracking models, and graph-based crowd analysis.
• Automated Industrial Inspections Enterprise-wide.
  • Scaled the system to 30+ plants, cutting costs by more than $10M annually and saving 30K labour hours.
  • Designed and deployed PISARES SIMS, an AI-powered drone and OCR-based inspection system.
  • Developed AI models for defect detection (corrosion, cracks, etc.) and engineering drawing digitization.
  • Mentored engineers in implementing AI models using PyTorch, ensuring sustainable AI adoption.
• Optimised AI for Transport Safety
  • Reduced AI model size from 1GB to 50MB (20x compression) while maintaining accuracy and transferring knowledge to the team.
  • Enabled real-time in-vehicle deployment across hundreds of Petronas’ fleet vehicles.

Director of Data Science

• Built and mentored a data science team, establishing best practices for AI-driven decision-making.
• Led AI and ML initiatives, including clustering, regression, and classification models.
• Designed and implemented an analytics platform integrating internal and external data for real-time insights.

Machine Learning Engineer

• Built feature comparing 10M products to extract matches in real-time by developing Deep Learning models.
• Built and deployed APIs using Flask to enhance analytic dashboard functionalities.
• Optimised ML pipelines for scalability, improving performance in large-scale data processing.

Machine Learning Engineer

• Developed an AI-driven trading robot, integrating ML strategies with broker APIs.
• Designed and implemented a risk evaluation system for financial companies across Asia using Python.

FC Otkritie

• Developed and implemented scoring models using SAS and R.
• Developed portfolio management strategies based on scoring models.

Graduate Certificate in Artificial Intelligence

• Natural Language Processing with Deep Learning
• Deep Learning for Computer Vision
• Machine Learning with Graphs
• Deep Generative Models
• Deep Multi-Task and Meta Learning
• Probabilistic Graphical Models

Bachelor's Degree

• GPA is 3.96

Course Facilitator for Professional Program in AI

• Facilitated Stanford’s “Machine Learning with Graphs” course (XCS224W), supporting professionals globally with personalized technical instruction through the AI Professional Program.

• Two times prize-winner of the federal districts stage competed in mathematics (high school)