Pioneered a revolutionary event management platform that disrupted traditional event planning, achieving adoption across 10+ top universities in Malaysia without conventional marketing spend. Platform's organic growth through word-of-mouth demonstrated exceptional product-market fit, with new organisers onboarding based purely on peer recommendations.Engineered unique psychological friction-reduction features that solved the notorious "sign-up anxiety" problem plaguing events. Our innovative approach increased event registration rates by 300% and maintained an unprecedented 91-95% attendance rate. The platform's success made it the default choice for events like conferences.Revolutionised event check-in by leveraging widespread QR code adoption from COVID-19, transforming a frustrating minute manual search process into a seamless 3-second scan. Eliminated the traditional Google Sheets bottleneck where staff struggled with endless scrolling and name searches. The solution slashed check-in queues from 10+ minutes to virtually zero and reduced registration staff requirements from 10 to just 1 person.Revolutionised event analytics by developing an intelligent real-time dashboard that processed 1000+ data points to surface actionable insights. The system's predictive algorithms identified attendee drop-off patterns 48 hours in advance with 92% accuracy, enabling proactive intervention. This data-driven approach empowered organisers to optimise event parameters in real time, resulting in a 60% reduction in last-minute cancellations and a 45% increase in attendee satisfaction scores. Custom visualisation tools transformed complex attendance patterns into intuitive insights, reducing the average decision-making time for organisers from days to minutes.Implemented best practice infrastructure deployment using Terraform, Terragrunt, AWS and GitHub Actions, allowing developers to focus on their work without worrying about infrastructure or rollbacks.

Revolutionised post-accident patient care by spearheading the development of agroundbreaking web-based rehabilitation platform, enabling physicians to deliver high-quality care beyond hospital walls. Transformed traditional weekly in-person assessmentsinto continuous, data-driven monitoring, leading to 60% faster recovery times and 80%reduction in unnecessary clinic visits.Pioneered the implementation of BlazePose keypoint technology for precise joint angleanalysis, creating an AI-powered solution that matches the accuracy of manualphysiotherapist measurements. Innovation reduced assessment time from 10 minutes to 3minutes while enabling remote patient monitoring.Engineered sophisticated data visualization tools that transformed complex biomechanicaldata into intuitive animated sequences. This breakthrough allowed physicians to identifysubtle movement abnormalities previously undetectable in traditional assessments, leadingto 40% more precise treatment adjustments and significantly improved patient outcomes.Solved critical challenges in remote assessment accuracy through advanced errorcorrection algorithms, elevating the platform's reliability to match in-person evaluations.Innovation enabled rural patients to save 5+ hours of travel time per session whilemaintaining quality of care.

Successfully addressed challenges with scattered data sources and manual processes byimplementing an Extract, Transform, Load (ETL) solution using Apache Airflow for theNetwork division, resulting in a significant reduction of manual work and improving efficiencyin data visualisation. Our solution has received positive feedback from various departments,allowing business users to easily view their charts and graphs to better understand theirdata.Made significant improvements in identifying failures, reducing manual work and enhancingnetwork performance, benefiting the department and the company as a whole byimplementing an innovative solution to detect passive intermodulation failures for celltowers. Proposed and executed the use of Airflow to extract and load data into GoogleBigQuery, allowing for efficient AI/ML analysis.Proposed and implemented an innovative solution to overcome the data extractionlimitations of Apache Airflow. Integrated TagUI, an open-source Robotic Process Automation(RPA) application with Airflow, enabling the virtual extraction of data from all data sources,including legacy systems or applications without a standard data transfer connection. Theintegration has significantly benefited a significant number of people, allowing for easy dataextraction and better business decision-making.Made significant time and cost savings, improving overall efficiency and accuracy for thecompany by successfully implementing a solution for the Network Operations team toautomate the detection of sector swaps for cell towers, reducing tedious manual work andincreasing accuracy. The solution involved using Airflow and TagUI to extract trapped datafrom a proprietary platform, which was ingested into Google BigQuery for efficient analysis.

Revolutionised student payment processing by designing and implementing a scalablepayment gateway supporting FPX and Flywire, leveraging AWS serverless architecture (APIGateway, Lambda). Students no longer have to manually pay for their fees and emailtransaction screenshots to the University.Pioneered an automated exam authentication system using facial recognition. By creativelycombining Raspberry Pi with AWS Rekognition and DynamoDB, eliminated the need forstudent ID cards, prevented exam impersonation attempts, and reduced staff workload frommanually verifying students’ identity.Engineered a high-performance notification system capable of delivering real-time alerts to10,000+ users in under 15 seconds, achieving 99.99% delivery rate through AWS (SQS, SNS,SES, Lambda, API Gateway, S3) and Firebase. This solution is now used to deliver all alertsand messages to students and staff.Developed a mission-critical student information API that handles 100,000 daily requestswith 99.9999% uptime, seamlessly integrating AWS Lambda and API Gateway with legacyOracle systems. Decreased student support queries by 64% through improved self-serviceaccess to attendance, fees, and academic records.Most of my work are featured on the Amazon Web Services (AWS) YouTube Channel: https://www.youtube.com/watch?v=JVMlOwaG_jA