Praxis #005 (ft. GovTech AI Practice)

​🌟Join us on Friday, for edition 5 of Praxis - dedicated sessions for technical deep dives and capabilities transfer.

​More details coming your way soon - keep an eye out! 👀

• ​Thiru (Data Scientist, GovTech’s AI Practice), will share the journey behind developing an AI agent for evaluating grants at Enterprise Singapore. His talk will cover the business problem that inspired the solution, the early approaches that failed, and the many iterations that eventually led to the creation of Weft—a flexible workflow engine that powers the current system, as well as the final overall solution. Weft is a powerful, asynchronous workflow orchestrator designed for Large Language Models (LLMs). It supports complex, multi-document workflows using a Directed Acyclic Graph (DAG) architecture. This enables users to design layered, sequential tasks—including conditional logic for document information extraction—while ensuring transparency, speed, and flexibility. Notably, Weft can handle both LLM-driven and traditional tasks, making it a versatile tool for AI-enabled automation.
  

• ​Tan Si Hao (Data Scientist, GovTech AI Practice) will be sharing on “LLM workflows as directed graphs - optimizations & ergonomics” (this is a related presentation to Thiru’s presentation on that same day, I can go after him)
  

• ​Pye Sone Kyaw (Data Scientist, GovTech) from GovTech's AI Practice will share on some of the learnings from the recent Nvidia GTC 2025 trip. The sharing will cover some of the more notable and interesting takeaways from the conference that can be applied to the work being done not just in AI Practice, but also in the wider AI space in general.

​About the speakers

• ​Thiru is a data scientist by training but don’t let the title fool you. He's spent most of his career in an identity crisis, building products to solve problems rather than creating machine learning models. His passion for tech-driven problem-solving has taken him through a whirlwind of data science, software engineering, and product adventures, including creating a digital map to help digitally disadvantaged hawkers during COVID-19.He holds a Master’s in Computer Science from the University of Edinburgh and a Bachelor’s in Business Analytics from NUS, where he was Valedictorian, School of Computing despite starting out in the Normal Academic stream. When he’s not busy turning caffeine into semi-functional code, he’s probably debugging life—one stack trace at a time.
  

• ​Tan Si Hao is a DS/SWE that works across the entire data platform stack.

  • ​

• ​Pye Sone Kyaw is a Data Scientist at GovTech's AI Practice where he leads the Multimodal AI track looking at all things multimodal AI and vision related, focusing on content authentication, generation, and insights.

​Note: Session is exclusive to Lorong AI members only!

​-----------
About Lorong AI

​​​Lorong AI is a co-working hub where AI practitioners connect, share knowledge, and grow through curated programming and a collaborative environment. Home to programmes like AI Wednesdays and AI ToolsDays, Lorong AI offers hands-on workshops, technical deep dives, and opportunities to solve real-world challenges with AI.