Experience

• 

  180 Degrees Consulting Singapore (Nanyang)

  • Singapore

  • Associate Consultant

    • Jan 2023 - May 2023
    • Singapore

    Participated in a consulting project for a Non-Profit Organization (NPO) in Singapore, which provides mentorship programs for youths in college. One of the primary responsibilities in this role included overseeing and executing copywriting tasks for various platforms, notably emails and LinkedIn.Demonstrated expertise in advising clients on content formatting strategies, for improved email click-through rates and user engagement on LinkedIn. The role required a deep understanding of research insights on the optimization of information distribution across diverse media formats, and its application to content creation.Further duties included recommending and utilizing public tools to streamline the content generation process. These recommendations, focusing on efficiency, proved instrumental in saving time and cost, transforming the pipeline of content creation for the organization.

• 

  Nanyang Technological University Students'​ Union (NTUSU)

  • Singapore

  • ITC Frontend Developer

    • Nov 2022 - May 2023
    • Singapore

    As part of the NTU Students' Union Information Technology Committee, contributed to restructuring the NTUSU portal's user experience and interface. The primary responsibilities involved developing features and troubleshooting to ensure optimal functionality.1. Frontend Feature Development & UI/UX Enhancement: Implemented a revamped campus facility reservation calendar page using Typescript on a React project; integrated the data delivered from the backend server, enabling a real-time view of available facility slots and respective timings. Enhanced the usability and improved the display of available reservation slots for respective dates and venues.2. Backend Feature Development: Developed a feedback form for the NTUSU portal using the Django framework, allowing users to report bugs and provide additional feedback.3. Collaborative Programming: Worked in a team environment of over 10 individuals, gained proficiency in managing compatibility issues and learned version control. This ensured the codebase functioned seamlessly across various environments, browsers, and device settings, demonstrating strong collaboration and communication skills.The role provided a unique platform to enhance technical skills, develop collaborative proficiency, and contribute meaningfully to the user experience of the NTU Students' Union portal.

• 

  Machine Learning and Data Analytics Lab at NTU EEE

  • Singapore

  • Computer Vision Research Engineer

    • Oct 2022 - May 2023
    • Singapore

    Modified the GazeML (Park et al.) Tensorflow model for real-time gaze estimation from a single RGB-camera input to achieve faster processing. Reconstructed the face detection, eye segmentation, and gaze vector visualization functions; resulted in a delay reduction of 94.33% in video output, from 3 seconds to 0.2 seconds. Additionally, the replacement of the Dlib face detection model to Google's mediapipe API allowed an accurate face detection and eye segmentation even in low-light conditions.Trained a PyTorch implementation of the model using a custom UnityEyes dataset and performed cross-dataset evaluations on the MPIIGaze dataset, ensuring model robustness in real-world applications. The trained model delivered a mean angular loss lower than 10 degrees. Alongside model training and evaluation, contributed to a comprehensive literature review, facilitating a broad understanding of the current gaze estimation research landscape. Assisted in model execution across diverse operating systems and development environments, enhancing the team's collaborative efficiency. The collective efforts in this role not only fostered technical growth but also highlighted the importance of teamwork in a research setting.

• 

  Deus Platforms

  • Singapore, Hong Kong

  • Project Lead, AI Specialist Backend Engineer, Business Developer

    • Apr 2022 - May 2023
    • Singapore, Hong Kong

    Contributions were made to the development of IO, an AI-driven platform redefining the e-learning experience. IO aims to create efficient, groundbreaking learning experiences, reducing task loads for educators and students.The project's highlight was the final technical demo - a web app utilizing large language models (LLMs), including OpenAI's GPT-4, to automate the creation of diverse educational content. Users supply essential structure and reference materials, enabling the system to auto-generate lecture notes, summaries, and question sets.LangChain was utilized to develop custom language models for specific tasks and integrated the PineCone vector database for efficient data search and retrieval with low latency. This ensured processing of queries by language models, regardless of the token volume in the original document.Faced with the challenges inherent in the current e-learning system, a key focus was the introduction of the IO platform as an effective solution. A scalable business model was proposed to confront these issues directly. SWOT analysis of market competitors laid the groundwork for the strategic planning for IO. This strategy has been designed to secure a steady expansion pathway, starting from initial deployment within selected universities. The long-term vision targets the broader online learning landscape and the customized AI assistant markets in the US, Southeast Asia, and Hong Kong.Achievements of the project include a nomination for the Esquel Group's Best Sustainable Idea Award at hackUST2022, receipt of HKD 50,000 in pre-seed funding, and ranking among the top 30 teams in the HKUST-Sino Entrepreneurship Competition 2023.