The One About Multimodal AI (Round V)
How is AI transforming the way we process, understand, and work with different types of data? Join us this episode to explore the field of multimodal AI - from practical applications in construction to the future of global language accessibility.
Ethan, (co-founder, Wenti Labs) will showcase one of their use cases that helps the construction team steamline their documentation process using an AI agent that is accessed via Telegram. They will share how they utilised OpenAI's SDK to build agents that could do function calling, file search, and image capability so that the agents can fit into their customers existing workflows without the limitation of legacy systems. Part of their workflow is a RAG pipeline that helps the team to retrieve the right clauses from BCA's Accessibility Code. (Technical Level: 200)
Jeremy Yee will be sharing more about the evolution of multimodal AI frameworks, from LLaVA to current innovations. Discover how different components work together to enable AI systems to understand and process multiple types of data - from text and images to video content. (Technical Level: 200)
Una Wang (Founder and CEO, LingoAI) and Henry Wang (Co-Founder, LingoAI) will share more about building a truly inclusive and decentralized multilingual AI economy at LingoAI. As AI systems become increasingly central to how we access information, interact, and transact globally, language remains one of the most significant barriers to participation. Drawing from real-world experience across Southeast Asia and Africa, Una will explore how LingoAI is enabling communities to not only contribute voice and language data but also reclaim ownership of it through blockchain and Web3.0 infrastructure. She will introduce the concept of LanguageDAOs, data monetization protocols, and how multilingual AI agents can serve as public infrastructure for education, local services, and digital economies. (Technical Level: 100)
More About the Speakers
Ethan started his career as a construction project manager with CapitaLand. Currently, at Wenti Labs, he's on a mission to vibe code solutions to help automate the mundane and repetitive workflows he used to complete in the construction sector. He has experience working at early-stage startups, tech companies as well as large enterprises.
Jeremy Yee is an Industrial Doctorate student at the SIT NVIDIA AI Centre, research interests include large language models and large multimodal models with a specific interest in Video as a modality. Jeremy holds an MSc in Scientific Computing and Data Analysis from Durham University and brings both academic rigor and practical insights to his work in AI development.
Una Wang is the Founder and CEO of LingoAI and the Founding Committee Member of Singapore Internet Governance Forum, brings a wealth of expertise in smart product design, virtual reality technology, and Web3-based AI innovation. Holding a Data Privacy and Technology certification from Harvard University, Una has successfully led the development of virtual reality applications and intelligent products. She envisions democratizing AI and Web3 technology, ensuring that data ownership and multiple languages AI accessibility are distributed equitably across the world. Her mission is to reshape the AI landscape through Web3, ensuring that AI development is inclusive, decentralized, and beneficial to all communities—not just a privileged few.
Henry Wang is the Cofounder of LingoAI, MetaLife, SmartMesh Foundation, Founding Committee Member of Singapore Internet Governance Forum (SG IGF) and the World Web3 Alliance (W3A), Henry Wang is a global thought leader and technology expert. His mission is “Connect the Unconnected, Bank the Unbanked”.
Want to be part of our sharing sessions? Sign up as a speaker here!