Singapore, Singapore
• Collaborated with a Nasdaq-listed mobile application development platform provider, extending on the platform by building a real-world Minimum Viable Model of an AI assistant co-pilot with autopilot capability through a multi-agent Large Language Model (LLM) framework interfacing with Microsoft Azure OpenAI Service • Engineered an automated AI program using multi-agent framework MetaGPT for generating full-scale mobile apps with minor human intervention geared towards no-code platform, producing first-cut program and optimising workflow • Delved into MetaGPT’s codebase to deeply understand framework with more than 50000 lines of code due to inadequate documentation, authoring sample code blueprints that demonstrated core functionalities such as agent initiation, action creation and inter-agent communication • Extended a GPT-powered mock backend server built in Flask by establishing connectivity to Microsoft Dynamics 365 Business Central, enabling retrieval and manipulation of real-world business data through dynamically generated webpages from the AI pipeline • Experimented with various Generative AI LLM frameworks such as LangChain, LangGraph, CrewAI and MetaGPT • Participated in agile development processes, contributing to Scrum-like activities such as sprints, backlog management and daily standups throughout a 7-month project timeline • Completed an intensive 2-month deep-skilling exercise spanning multiple topics such as Machine Learning Operations (MLOps), Computer Vision, Time Series Analysis, Natural Language Processing (NLP) and Large Language Models (LLMs)
• Sole backend developer for urban delivery optimisation project, leading system integration and backend development efforts • Developed RESTful APIs to interface with MySQL database and route optimiser AI engine, ensuring seamless communication between ReactJS frontend and Java backend services • Formulated unit test cases for the created APIs using JUnit to ensure code quality • Configured Apache Tomcat servers for deploying and hosting optimiser engine on virtual machines, leveraging SSH for remote server management