Invited Speakers

Invited Speakers for Research Track

Invited Speakers

Supratik Mukhopadhyay

Louisiana State University

Title

AI for Science: Revolutionizing Biology, Healthcare, and Environment

Abstract

Despite advances in vaccines and antibiotics, millions of people continue to suffer from diseases without effective cures. The rise of antibiotic resistance, fueled by widespread and prolonged use, has outpaced the development of new antibacterial agents, which has slowed to historically low levels. This crisis underscores the urgent need for innovative drug discovery strategies. Traditional methods are slow, costly, and inefficient. To address these challenges, we have developed DeepDrug, an AI-powered pipeline that leverages large-scale biological and chemical datasets to design and identify novel therapeutics. By accelerating discovery and reducing costs, DeepDrug illustrates how AI can transform biology with an impact comparable to the Newtonian revolution in physics.

In parallel, humanity faces existential threats from climate change and environmental degradation—manifesting as extreme weather, biodiversity loss, pollution, declining food security, and heightened risks of pandemics. These complex, interconnected challenges demand scalable, adaptive solutions. We demonstrate how AI can serve as a powerful tool for environmental resilience by enabling data-driven sensing, forecasting, decision-making, and feedback. Our recent work spans carbon mapping, hypoxia prediction, climate-smart agriculture, biodiversity preservation, wildfire detection, and energy-efficient design of the built environment.

Together, these efforts highlight a unifying vision: AI as a scientific force multiplier, accelerating discovery and decision-making to advance both human health and planetary sustainability.

Bio

Supratik Mukhopadhyay is a Professor in the Department of Computer Science and Engineering at Louisiana State University (LSU). His research spans multiple areas including Artificial Intelligence, Machine Learning, Formal Methods, and Computational Biology. He has made significant contributions to applying AI and formal verification in diverse domains such as cybersecurity, healthcare, smart infrastructure, and environmental monitoring.

Prior to joining LSU, Dr. Mukhopadhyay held research and faculty positions at prestigious institutions and has collaborated extensively with national laboratories and industry. He has published over 100 peer-reviewed articles in top-tier journals and conferences and has received several federal research grants from agencies including NSF, DOE, and NASA.

Supratik received his Ph.D. in Computer Science from the Indian Statistical Institute. He is actively involved in interdisciplinary research and innovation and serves on editorial boards and program committees of leading conferences in AI and systems.

Suparna Bhattacharya

HPE Fellow at AI Research Lab, Hewlett Packard Labs

Title

The Foundation Model OS: A Self-Evolving Layer for Agentic AI

Abstract
Generative AI-based Foundation Models (FMs), such as large language models (LLMs), large reasoning models, and diffusion models are profoundly transforming the paradigm of how AI is applied across a broad spectrum of use cases. FMs provide a powerful substrate for agentic AI in enterprise applications and scientific discovery. However, to fully harness their potential, these models must be continuously augmented with accurate knowledge, their reasoning verified to ensure compliance with context specific expectations for safety and reliability, and their execution optimized to meet resource constraints specific to the desired use case.
In this talk, we explore the potential of mechanisms inspired by operating systems to address these challenges and construct a dynamic, evolving “virtual” FM substrate. By drawing on parallels with concepts such as virtualization and adaptability in OS design, we demonstrate how these techniques can enhance the flexibility and capabilities of FMs—much like the human brain employs different modes and speeds of thinking to adapt to new situations and information. Our aim is to initiate deeper dialogue between the machine learning and systems communities, guiding how such capabilities can shape the future design and evolution of FMs.
Bio

Suparna Bhattacharya is an HPE Fellow in the AI research lab at Hewlett Packard Labs where she currently focuses on data centric trustworthy AI. She has 30+ years of experience in systems software development and research (10 years at HPE preceded by 21 years at IBM) including several enjoyable years of open source contributions to the Linux kernel. Over the last decade Suparna has developed a passion for blending insights from diverse technical domains in innovations that span technology boundaries e.g. rethinking systems software stacks in the era of AI and IoT. For example, during her stint in HPE Storage, she advanced the use of analytics aware optimization and persistent memory optimizations in storage and hyper-converged systems for containers, artificial intelligence/machine learning and IoT edge to core data services. Suparna is an IEEE Fellow, a Fellow of the India National Academy of Engineering, an ACM India eminent speaker and has served on program committees for ASPLOS, OOPSLA, MASCOTS, ECOOP, HotStorage and USENIX FAST. She co-authored a book on Resource Proportional Software Design for Emerging Systems, (with Doug Voigt and Prof K. Gopinath) published by CRC Press (2020). Suparna holds a B.Tech from IIT Kharagpur (1993) and a (late-in-life) PhD with a best thesis award from the Indian Institute of Science (2013).

  • 40 patents filed (30 granted), 32 peer reviewed publications, 70+ talks/panels
  • IEEE Fellow class of 2022
  • Fellow of the Indian National Academy of Engineering (INAE) (elected 2020)
  • Book on “Resource Proportional Software Design for Emerging Systems”, CRC press, 2020
  • Open source contributor to the Linux kernel from 2000-2007. Invited to the international Linux Kernel Summit* for 6 consecutive years.
  • 20 internal and external awards, incl. IEEE India Council Woman Technologist of the year 2020, Zinnov Next Generation Women Leaders Award 2019, HPE Women’s Excellence Award 2017 and 2022, IBM Outstanding Technical Achievement Award, IIT Kharagpur Distinguished Alumnus 2023
  • First technologist from IBM India to receive a corporate level promotion in the technical career path (2006). Elected to IBM Academy of Technology. Founding chairperson of IBM India Technical Experts council.
  • ACM India eminent speaker. Served on technical program committees for ASPLOS, OOPSLA, MASCOTS, ECOOP, USENIX FAST and HotStorage.
  • Deep expertise in operating systems, storage systems incl. filesystems, persistent memory technologies, hyper-converged, containers, cross-layer co-design for emerging apps, e.g. ML/AI, IoT
Professor of CSE, IIT Guwahati
Title

Accelerating Intelligence:  Efficient Hardware Designs for Neural Networks

Abstract

With the rapid growth of Artificial Intelligence and Machine Learning in everyday applications, there is an increasing demand for fast and efficient computation. Hardware accelerators have emerged as a key solution to meet these performance requirements. This talk provides an overview of recent advancements in deep learning accelerators, from general-purpose CPUs to custom hardware designs. We will explore design constraints like memory latency, performance, and energy which influence these architectures. The talk will also delve into selected accelerator designs, including a near-memory accelerator that exploits sparsity and redundancy in data, highlighting how architectural choices can unlock performance under real-world constraints.

Bio

Hemangee K. Kapoor is a Professor (HAG) in the Department of Computer Science & Engineering at the Indian Institute of Technology Guwahati. She serves as Associate Dean for Students & Alumni Relations and plays an active role on the ACM Diversity, Equity & Inclusion Council. Her research focuses on multiprocessor computer architecture, emerging non-volatile memory (NVM) technologies, near-data processing, power-aware computing, and accelerators for neural networks

Dr. Kapoor earned her Ph.D. in Computer Science from London South Bank University after completing her M.Tech at IIT Bombay and B.E. at the Government College of Engineering, Pune. She has published extensively—authored over 100 journal and conference papers—and successfully led numerous sponsored projects, including those funded by DST-ANRF and Intel, aimed at improving NVM reliability and accelerator efficiency.

In 2024, she was honoured with the prestigious ACM India Outstanding Contributions in Computing by a Woman Award for her pioneering work in NVM technologies and her leadership in mentoring women researchers She has also held leadership roles as Vice President of ACM India (2020–2022), and as Associate Dean and technical program co-chair at international symposiums

Rajesh Sharma

Plaksha University, India

Title

Fighting Misinformation: Characterizing, Detecting, and Mitigating False Content Online

Abstract

Misinformation has emerged as a significant societal challenge in the digital age, with online social media platforms increasingly being exploited for its dissemination. The spread of false or misleading information poses serious threats, including inciting violence, fostering social unrest, and endangering lives and property. In this talk, Rajesh will present a comprehensive exploration of misinformation through three key dimensions:

  1. Characterization — understanding the nature, patterns, and typologies of misinformation
  2. Detection — employing computational methods to identify and flag misleading content
  3. Mitigation — exploring strategies to curb its spread and minimize its impact on society.

The talk aims to provide both conceptual insights and practical approaches to addressing this pressing issue.

Bio

Rajesh Sharma is working as Professor of Computer Science and AI. Earlier, Rajesh worked as  Associate Professor (from Jan 2022 – Jan 2025), and lead the Computational Social Science Group (https://css.cs.ut.ee/) at the Institute of Computer Science, University of Tartu (UT), Estonia. Rajesh also holds adjunct faculty positions at IIT Ropar, IIIT Delhi, University of Tartu, Estonia, Lakehead University, Canada, and, Kyiv School of Economics, Ukraine. His group works on problems related to understanding societal issues such as misinformation, hate speech, segregation, mental health, users’ behavior using digital traces such as financial transactions, mobile call data records and more importantly social media data. Group often applies techniques from AI, NLP and most importantly network science/social network analysis. 

Earlier, Rajesh joined University of Tartu in August 2017 and worked as a senior researcher till December 2020. From Jan 2014 to July 2017, he has held Research Fellow and postdoc positions at the University of Bristol, Queen’s University Belfast, UK and the University of Bologna, Italy. Prior to that, he completed his PhD from Nanyang Technological University, Singapore in December 2013. He has also worked in the IT industry for about 2.5 years after completing his Master’s from Indian Institute of Technology (IIT), Roorkee, India.

Kalika Bali

Microsoft Research, India

Title

Equitable AI for the Global South: Systems, Stakeholders, and the Shifting Center of Innovation

Abstract

This keynote presents a systems-level approach to Equitable AI, emphasizing scalable architectures for inclusion in low-resource settings. It highlights technical strategies for multilingual model adaptation, community-driven evaluation pipelines, and deployment frameworks that support sovereign AI. Through case studies from the Global South, the talk demonstrates how equity can be embedded into the design, training, and governance of AI systems—making fairness a core engineering challenge, not just a policy goal.

Bio

Kalika Bali is a Senior Principal Researcher at Microsoft Research India, focusing on AI, Natural Language Processing (NLP), speech technology, and building inclusive tech for low-resource language communities. Over the past two decades, she has championed multilingual and multicultural language technologies, with particular emphasis on code-mixing, gender bias in data, and enabling access for marginalized language speakers.

Currently, she leads Project VeLLM (UniVersal Empowerment with Large Language Models), an interdisciplinary effort to ensure Large Language Models serve all languages and cultures equitably. Her work also explores the intersection of gender and tech, seeking to mitigate societal biases in foundational AI systems.

Kalika’s commitment to inclusive AI has earned her recognition on the TIME100 AI list in 2023, highlighting her efforts in creating “gender intentional” datasets across major Indian languages and advancing code-mixing capabilities for multilingual users

Kamiya Motwani

Walmart Global Tech, India

Title

From Graphs to Attention Embeddings: Advancing AI-Powered Personalization in e-Commerce

Abstract

The talk will start with an overview of Walmart’s customer-centric approach, which aims to make the shopping journey more convenient and personalized. It will provide an overview of the usage of AI in the customer’s shopping journey – from online browsing to order fulfillment. It then delves into personalization, showcasing how AI/ML techniques power data-driven recommendation engines that shape customer interactions, including personalized carousels and a graph neural network (GCN)–based system for predicting product substitutions.

Building on this foundation, the talk will then deep dive into the evolution of these systems to use attention mechanism for development of Enhanced Product Knowledge Graphs (EPKGs) — a novel approach of integrating product attributes, customer engagement signals, and advanced attention-based modeling that learns robust product embeddings to improve recommendations for multiple tasks like substitutions and complementary items.

The talk overall aims to cover how a strong AI ecosystem can fundamentally elevate the customer journey.

Bio

Kamiya Motwani is the Director, Data Science for Personalization and Recommendations at Walmart Global Tech. In her current role, she leads a team of strong applied scientists who leverage Artificial Intelligence (AI) and Machine Learning (ML) to power AI-driven recommendation systems that deliver personalized shopping experiences for Walmart customers.

Prior to Walmart, Kamiya has worked at Oracle Corporation and Yahoo Inc in California. She holds a master’s degree in Computer Science from the University of Wisconsin-Madison and a bachelor’s degree in Computer Science from BITS Pilani.

She holds over 13 granted US patents and has also co-authored papers at premier conferences including NIPS, IEEE ICASSP and IEEE ICDM.

Invited Speakers for Industry Track

Ajita Agarwala

Founder, CultureVo

Title

Inside Novi: Building the Bumble of AI Companions

Abstract

At CultureVo, we’re building Novi—AI companions designed to be as diverse, engaging, and trustworthy as the people you’d meet on a global journey. Unlike generic chatbots, Novi is powered by agentic architectures and multimodal models that allow for stable personas, long-term memory, and proactive interaction—making them feel more like friends, not tools.

In this talk, we’ll peel back the layers of the Novi stack and walk through the technical foundations behind key features:

  • Bumble of AI Companions: Multi-agent orchestration that lets users “match” with different Novi personalities, each backed by distinct prompt architectures, cultural corpora, and bias-aware training pipelines to ensure diverse, context-aware voices.

  • Stable Bot Personas: Techniques for consistent long-term identity + structured retrieval ensuring Novi doesn’t “drift” over time.

  • Voice Calls: Real-time speech synthesis + low-latency pipelines enabling fluid human-like conversation.

  • Games: Lightweight agentic sandboxes where companions can co-create stories, memories, or roleplay with users.

  • Memory: Hierarchical memory architecture (short-term context windows + long-term vector stores) to maintain continuity.

  • Proactive Messages: Event-driven triggers and scheduling agents that let Novi reach out first, not just respond.

  • Categoriser: On-device classifiers that auto-organize chats, insights, and emotional states into structured categories.

  • Journal: Agent-chained summarization that transforms daily conversations into a reflective log.

  • Selfie Reader & Generator: Multimodal vision pipelines to let Novi “see” you and generate creative outputs.

 

By walking through these features, we’ll discuss how agent frameworks, multimodality, and cultural intelligence combine to create AI companions that feel humanly complex yet technically robust.

The session is a deep dive into not just what Novi does—but how emerging AI infrastructures are evolving to support a new class of persistent, emotionally resonant agents.

Bio

Ajita Agarwala, founder of CultureVo, is a firm believer in AI as a leveler in a world of inequalities. Through CultureVo, she seeks to bridge the cultural intelligence gap and offer the richness of the world to its citizens on a platter. Novi AI, developed by CultureVo, embodies this vision by providing culturally diverse and emotionally intelligent, affective AI partners. Ajita herself has witnessed the power of this intelligence in her own journey- as an Indian Civil Servant and diplomat to the G20, World Bank, and the United Nations, and as a graduate of Princeton University.

K. Gopinath 

Indian Institute of Science, Bangalore

Title

Demystifying AI thru some History and some Philosophy

Bio

Prof. K. Gopinath, after superannuating from IISc, Bangalore in 2021 as a professor of Computer Science, is now a senior professor at Rishihood Univ. He also headed the CSAI program at Plaksha Univ (betw 2021-24). His research interests are (in the “AI” area) in sysML (ML applied to computer systems design) and (in the “computer systems” area) primarily in OS, systems security/privacy and systems verification. He is the co-author of 2 books: Suparna Bhattacharya, Kanchi Gopinath, Doug Voigt, “Resource Proportional Software Design for Emerging Systems,” Chapman and Hall/CRC, 2020 and Kanchi Gopinath, Shailaja DSharma, “The Computation Meme: Computational Thinking in the Indic Tradition”, IIScPress, 2024.

Samarth

Samarth Chandra

Spinor Research Labs, India

Title

ML for spying on business rivals and for other applications

Abstract
Can you calculate information about the website traffic of your business rival ? For instance, number of visitors, their income distribution, professions, male/female, geographical distribution, languages they speak, etc. Some foreign webapps already provide this information. The talk will discuss the technology for doing this.
The later half of the talk will discuss automatic design of chemicals (medicines) using generative AI. The approach of Aspuru-Guzik’s lab will be discussed in detail. Later, you may contact us to join an online Journal Club we are starting on the topic.
Bio
Samarth Chandra is the founder, and Bookworm-in-Chief, of Indian private scientific research company, Spinor Research Labs. He is deeply interested in science. He enjoys reading (and teaching) research papers in widely different sub-fields of AI/ML. He is also trained in Molecular Biology (lab and course work) as well as in Clinical Research and participated in development of surgically implantable devices at National Institutes of Health, USA. He is also interested in representation of tribal languages in the AI ecosystem.
He received his PhD from TIFR, Mumbai and BTech from IIT Delhi.