Singapore
I love everything about technology, intrigued by this ever growing industry, an ever eager team player that pushes for adaptability and innovation, and if need be, I would be grateful for any opportunity to lead and hone my skills. The field of AI, with its dynamic and rapidly changing nature, captivates me. It truly amazes me how it can incorporate into our daily lives to empower our routines so easily and seamlessly. I always believe in bringing that human touch of tech, trying to close the bridge between humans and machines, with AI democratisation and accessibility in today's world, the prospect of working on projects inclusive of AI is genuinely exciting and I look forward to embracing the challenges and opportunities it presents. In my free time, I also like to shoot photos on my camera!📷 Happy to Connect!
Worked on Full Stacked Development on an Enterprise Level Multi-Agent Agentic Orchestration Application, involving a Next.js frontend, middleware and backend, with CrewAI and FastAPI integrations. Wrote custom tools for Agents to utilise, pioneering in Multi-Agent Orchestration to enable full workflow automation, integrating traditional Robotic Process Automation(RPA) with Agentic AI Utilised Traditional Machine Learning model tuning methods with comprehensive data pre-processing and massaging with NumPy, Spark and Pandas for an XGboost Predictive model retraining, exploring MLOps for further integration of newer datasets. The XGboost model aids in predicting approvals for schemes within the board, increasing application approval times by 100% for officers of the board and impacting thousands of citizens applying for scheme approval
Imagery Support Group (ISG) Formation, Intelligence Analyst. Led a team of Programmers in creating a Telegram Bot for the Formation. This involved using Python and Telegram API and was deployed on Google Cloud. This multifunctional bot has user authentication using a designated PIN allocated to any authorised user. User Authorisation Control was also a big part of the project where different officer in charges had different access to functions. The bot acts primarily as a scheduling bot for users to schedule poignant events as Reminders for any text group, especially during the periodic COVID ART tests, where multiple reminders had to be sent out previously by Regulars. Additionally, it also interacts with an SAF indentation site using Selenium, enabling both NSFs and Regulars to indent ration through telegram, while it updates the site in the background with the user's choices. As a privacy feature, the telegram API enabled us to restrict the bot from reading messages but rather only send them out. During a time of Covid Restricted Measures, crucial data protected under the Personal Data Protection Act (PDPA) was not accessed by the bot.
Extended System Integration Testing and User Acceptance Testing along with Development Operations and Test Project Management
Systems Integration Testing and User Acceptance Testing for Site LicenceOne (Now GoBusiness Gov SG)