Singapore Meetup: Gen AI Frontiers [Resaro, Salesforce]

We are thrilled to announce our first in-person Meetup in 2024! Building upon the feedback from our last gathering in November, we're back to diving even deeper into the realm of Generative AI. This time, our speakers from Resaro and Salesforce will delve into practical and cutting-edge topics. Mark your calendars for the QuantumBlack Southeast Asia Meetup, happening on Wednesday, 24 April 2024, from 6:00 PM to 7:30 PM SGT / GMT+8, at McKinsey & Company Singapore Office.

***

AGENDA (SGT, GMT+8)
5:30 PM - 6:00 PM: Arrival and Networking
6:05 PM - 6:10 PM: Welcome & Introductions
6:10 PM – 6:30 PM: “Charting the Deepfake Evaluation Frontier: From Governance to Testing” by April Chin + Dr. Sreejith Balakrishnan @ Resaro
6:30 PM – 6:40 PM: Q&A
6:40 PM – 7:00 PM: “Foundation Models for Universal Time Series Forecasting” by Gerald Woo @ Salesforce Research Asia, Singapore Management University
7:00 PM – 7:10 PM: Q&A
7:10 PM – 7:15 PM: Closing address
7:15 PM onwards: Networking with refreshment

***

TOPICS:
“Charting the Deepfake Evaluation Frontier: From Governance to Testing” by April Chin + Dr. Sreejith Balakrishnan
In this talk, we will delve into the rapidly evolving domain of deepfake content governance and explore the intricacies of assessing deepfake detection technologies. We aim to compare and contrast the various regulatory frameworks that exist today in regards to how they shape the handling of deepfake media. We will then share the various challenges involved in evaluating the efficacy of deepfake detection solutions and offer insights into Resaro's comprehensive evaluation framework. Attendees will gain a nuanced understanding of the balance between technological advancements in content creation and the imperative measures being implemented to ensure digital integrity and trust in the multimedia landscape.

“Foundation Models for Universal Time Series Forecasting” by Gerald Woo
In the era of foundation models, deep learning for time series forecasting is facing a revolution. While the existing paradigm operates in a one-model-per-dataset setting, foundation models for time series forecasting collates massive amounts of time series data from diverse domains to build a single Large Time Series Model (LTM). Such a model should have strong zero-shot forecasting capabilities on unseen time series data. In this presentation, I will introduce Moirai, our recent foundation model for time series forecasting, and explain how we solved the various challenges of building such a model.

***

SPEAKERS:
April Chin, Managing Director and CEO (Singapore) @ Resaro
April Chin serves as the Managing Director and CEO (Singapore) at Resaro, an AI assurance company specialising in independent, third-party testing of mission-critical AI systems.
April's commitment to Responsible AI began in 2019 at BasisAI, where she played a pivotal role in driving the adoption of governance-by-deployment approaches in AI solutions among enterprises. She has also led multidisciplinary initiatives on the operationalisation of AI governance frameworks into executable processes for business and technology teams.
Drawing from over a decade of experience as a former Singapore Administrative Service officer, April has collaborated closely with senior government officials and C-suite leaders to shape future-ready policies that are anchored on governance principles. As a thought leader in this field, April shares her expertise as a Futures Fellow at the Lee Kuan Yew School of Public Policy and as an Advisor-Expert at the Chandler Institute of Governance.

Dr. Sreejith Balakrishnan, Senior Responsible AI Scientist @ Resaro
Dr. Sreejith Balakrishnan serves as a Senior Responsible AI Scientist at Resaro, where his work centers on creating protocols for testing the performance and robustness of AI models. In addition, he is a member of the AI Technical Committee under the Singapore Standards Council (SSC) and represents Singapore in the SC42 subcommittee of ISO, the International Organization for Standardization. Holding a PhD in Computer Science from the National University of Singapore (NUS), his research focus was in the fields of AI value alignment and algorithmic fairness, leading to publications in prominent conferences such as NeurIPS and AIES and earning him the 2020 NUS Research Achievement Award. Prior to his tenure at Resaro, Sreejith gained extensive experience at ABB Pte. Ltd., Singapore, leading a team dedicated to addressing complex issues in process automation.

Gerald Woo, Industrial PhD Candidate @ Salesforce Research Asia, Singapore Management University
Gerald Woo is a Ph.D. candidate in the Industrial PhD Program at Singapore Management University and a researcher at Salesforce AI Research Asia. His research focuses on deep learning for time-series, including representation learning, and forecasting.