Speakers at ADCOM 2024
Prof. Virginia Dignum
Umeå University, Sweden
Prof. Virginia Dignum
Umeå University, Sweden
Bio
Prof. Frank Dignum leads a research group in the field of socially conscious AI and is currently a Professor of Socially-Aware AI at Umeå University and an associate professor at the Department of Information and Computing Sciences of the Utrecht University. Dignum is best known from his work on software agents, multi-agent systems and fundamental aspects of social agents. Know more
Dr. Naveen Yeri
Wells Fargo
Abstract
Customers’ financial decisions are backed by reasons that analytical approaches may not comprehend. We sense it through their experiences and needs. It is in the emotions and trust we cultivate that customers find security in our services, not merely in rational explanations. This, then, is the essence of understanding customers through their emotions and trust, creating a foundation of reliability beyond mere transactional logic. In my talk, I expose the need to understand the emotions behind transactions, discover features that can provide tremendous explanatory power in analytics, and recommend a framework to leverage AI toward customers’ success. This unique framework can be extensible to current infrastructure, uses generative methods and blends in ethics and responsibility aspects.
Bio
Dr. Naveen Yeri heads the Enterprise Analytics and Data Science function at Wells Fargo. In this delivery role, he is responsible to orchestrate Artificial Intelligence and Generative AI products for the enterprise. He also leads the Business Intelligence Capability Center and champions the Analytics Community of Practice connecting with 5,000+ analytics professionals in the India and Philippines region.
Naveen has been in the financial services sector for 24 years in the US and India regions. In his prior roles, he led Data Science and Platform teams for Bank of America and IBM, serving Consumer and Small Business customers across Risk, Marketing and Technology groups.
He completed his Doctorate from Rennes School of Business, France in 2024, focusing on Emotion Intelligence and Contextual Insights using Complaints. He has a MS degree in Operations Research from Virginia Tech and MBA from UNC-Charlotte.
He is on the advisory board of Rennes School of Business, and an active speaker at industry and academic conferences. Naveen has 20 patents and over 50 pending applications, he passionately grooms professionals to be innovation mentors.
Dr. Sameep Mehta
IBM Research
Title – Building Responsible and Trusted AI Systems
Abstract
In this talk, we will cover various dimensions of Trusted AI. While building AI model in correct way is a significant part of Safe AI, we will articulate that trust needs to be infused in all aspects from Data to Model Building to Deployment. We will showcase some of the work done in the context of FMs. The talk will conclude with some open problems and call for community collaboration.
Bio
Sameep Mehta is IBM Distinguished Engineer in the area of AI at IBM Research AI. He also holds Adjunct Faculty Position at IIT Jodhpur and IIIT Delhi. His research interests are in Foundation Models, Governance, and Meta Data Management. He has co-authored the book “AI for You – the New Gamechanger” for adoption of AI by Enterprises. Prior to IBM Research, Sameep received his Masters and Ph.D. from The Ohio State University in 2005 and 2006 respectively.
Sakina Pitalwala
Intel
Title: Assisted chip-design in the era of Large Language
Abstract
Over the years, the chip design process has seen a surge of AI-powered applications that have optimized design workflows, reduced time-to-market, and streamlined engineering efforts. Concurrently, the remarkable advancements in large language models (LLMs) for natural language processing tasks have inspired researchers to explore their potential in adjacent domains which have similarities with natural language, like code. The impressive performance of LLMs in software-development tasks, including code generation, review, and debugging, has naturally led to investigations into their applicability to the hardware development domain as well. In this talk, we seek to shed light on the transformative potential of large language models in streamlining and enhancing the complex chip design process, ultimately contributing to the continued advancements in hardware development.
Divyashree Tummalapalli
Intel
Title: Assisted chip-design in the era of Large Language
Abstract
Over the years, the chip design process has seen a surge of AI-powered applications that have optimized design workflows, reduced time-to-market, and streamlined engineering efforts. Concurrently, the remarkable advancements in large language models (LLMs) for natural language processing tasks have inspired researchers to explore their potential in adjacent domains which have similarities with natural language, like code. The impressive performance of LLMs in software-development tasks, including code generation, review, and debugging, has naturally led to investigations into their applicability to the hardware development domain as well. In this talk, we seek to shed light on the transformative potential of large language models in streamlining and enhancing the complex chip design process, ultimately contributing to the continued advancements in hardware development.
Dr. Srinivasan Iyengar
Microsoft
Dr. Hari Bhaskar
Google
Title: Online Content Moderation at scale
Abstract
We heavily rely on the Web for meeting our information needs today. Examples include Wikipedia, Twitter, Instagram, Youtube, Google Maps etc. All of these are platforms where millions of users post billions of pieces of content every day on a wide range of topics. The content is consumed by hundreds of millions of users. While a rich source of information, these platforms are also easy targets for abuse and harm, both intentional as well as un-intentional. Intentional harm includes the use of these platforms for fraud, misinformation, trolling, hate and other forms of vandalism. Unintentional harm includes factually incorrect information, stale information, information bias and more.
The underlying platform providers have a huge responsibility in terms of ensuring that users are provided a safe, delightful, useful and transparent experience for the information that is presented to them. However, this is a very difficult problem to solve in the real world. There are a large number of very hard challenges to deal with.
For example:
How do we deal with highly motivated bad actors who are technically savvy, have financial means and put continuous efforts to identify vulnerabilities and expose them, adapting at a very fast pace?
How do we differentiate between facts and opinions and determine the actual “ground truth label” and do it at scale, with thorough representation for all kinds of patterns and do it fast before the patterns change so that our models have a good exposure to the underlying problem
How do we deal with signal sparsity? What if we simply don’t have any useful signals to train the classifier on?
What do we do if the tolerance for errors and mistakes is very low and yet, we don’t have an effective classifier to solve the problem?
How do we solve this for a large number of semantically different types of documents?
How do we build the system to deal with billions of documents and build an efficient and reliable system?
In this short talk, we will elaborate on this super critical and pressing challenge faced by the tech industry and give you insights into how the above challenges and many more are solved.
Bio
Hari is a multi-faceted engineering leader with a good blend of research and product engineering expertise. I have a research background with a specialisation in AI & ML. Hari is passionate about building AI and ML apps that can drive value to the end users and by large the society.
Hari has 13 patents granted by USPTO, 30+ publications in various national & international conferences by IEEE, Springer, Elsevier, Dpubs and institutes like Indian Institute of Sciences, Indian Statistical Institute, IIMB. The area of publications include machine learning, open government data, big data reference architectures. Hari is a writer, speaker and likes to share on technology, inspirational, and social topics.
Domain expertise: Banking, Finance, Enterprise Content Management, Healthcare, Travel, Maps & User generated content
Dr. P Kumaraguru
IIIT Hyderabad
Dr. Neelam Sinha
Indian Institute of Science
Title: Persistent Homology on fMRI times series to understand Mild Cognitive Impairment (MCI)
Abstract:
Persistent Homology is a branch of Topological Data Analysis (TDA) that examines the “shape of the data”, to arrive at inferences. The idea of “shape”, which is usually reserved for tangible structures, can also be used to describe functional brain networks. In this talk, we will look at applications of Persistent Homology for identifying brain network alterations that occur in a particular neurodegenerative disorder called Mild Cognitive impairment (MCI).
Bio:
Neelam Sinha is a faculty at CBR, Center for Brain Research. She was earlier a faculty at IIIT-Bangalore. At CBR she leads the Multi-modal Neuroimaging lab. Her group works on understanding neuro-degenerative disorders using MR images and EEG signals.