USF Data Science & Artificial Intelligence Speaker Series

We are excited to welcome Suman Debnath, Technical Lead in Machine Learning at Anyscale, for a practical and intuitive introduction to distributed training.

Talk Description:
As modern AI models continue to grow, single-GPU training is no longer enough. Distributed training has become essential, but scaling models introduces challenges that require understanding communication patterns, system bottlenecks, and key trade-offs.
In this session, we will break down distributed training from first principles. We will explore why single-GPU training hits limits, how transformer models manage memory, and what techniques like gradient accumulation, checkpointing, and data parallelism actually do.
We will also demystify communication primitives, walk through ZeRO-1, ZeRO-2, ZeRO-3 and FSDP, and show how compute and communication can be overlapped for better efficiency. Finally, we will connect these concepts to real-world tooling used in frameworks like Ray and PyTorch. Attendees will gain a clear, grounded understanding of how distributed training works and when to apply different strategies.

Bio:
Suman Debnath is a Technical Lead in Machine Learning at Anyscale, where he works on large-scale distributed training, fine-tuning, and inference optimization in the cloud. His expertise spans Natural Language Processing, Large Language Models, and Retrieval-Augmented Generation.
He has also spoken at more than one hundred global conferences and events, including PyCon, PyData, and ODSC, and has previously built performance benchmarking tools for distributed storage systems.

We look forward to seeing you!

#DataScience #MachineLearning #DistributedTraining #Ray #PyTorch #LLM #RAG #DeepLearning #USFCA #USFMSDSAI #DataInstitute #AIEngineering #TechTalk

We are excited to welcome Punn Kamolyabutr for an upcoming Data Science Speaker Series talk.

Punn is the CTO and co-founder of Conduit (Y Combinator W24), a platform for building conversational AI automations. Previously, he worked on AI at Google, built systems at a crypto lending protocol, and worked as a trader at a private wealth fund.

While many AI talks focus on technology, this session focuses on what AI enables: massive leverage for small teams. Punn will share how a small engineering team supports a platform sending millions of messages and hundreds of thousands of calls each week, how AI is reshaping team structures across engineering and non-engineering roles, and why leading companies are hiring fewer but sharper people.

This talk is especially relevant for students interested in startups, early-stage teams, and understanding how the technical job market is evolving.

Join us and be part of the discussion.

#USFCA #USFMSDSAI #DataInstitute #DataScience #ArtificialIntelligence #Startups #AIEngineering #TechTalk #StartupCareers #EarlyStageStartups

We’re thrilled to welcome back Josh Starmer back to the USF Data Science Speaker Series!

Description: Although Large Language Models and AI are known to generate false and misleading responses to prompts, relatively little effort has gone into understanding how we can quantify the confidence we should have in the output from these models. In this seminar, I will illustrate the problem using a simple neural network and then demonstrate two methods for quantifying our confidence in the model outputs. I will then show how these methods can be applied to Large Language Models and AI.

About the Speaker: Josh Starmer is the person behind the popular YouTube channel, “StatQuest with Josh Starmer.” Since 2016, Josh has used an innovative and unique visual style to clearly explain Statistics, Data Science and Machine Learning concepts and algorithms to curious people all over the world. Rather than dumb down the material, Josh brings people up with simple examples worked through, step-by-step, using pictures to make sure every main idea is easy to understand and remember. By breaking down even the most complicated algorithms into bite sized pieces, StatQuest has helped people, all over the world, win data science competitions, pass exams, graduate from school, and get jobs and promotions.

RSVP now to secure your spot!

#USFCA #USFMSDSAI #DataInstitute #DataScience #MachineLearning #ArtificialIntelligence #LLM #StatQuest #TechTalk

We are excited to welcome David Schoeller-Diaz, USF Class of 2006 alumnus and global program leader, for a special session on practical ethics in data science and artificial intelligence. This talk complements our cohort’s ethics module and the USF Data Science Speaker Series.

Talk Description: Most ethical decisions in data science are not dramatic dilemmas. Rather, they are embedded in everyday choices about how we formulate problems, what we include or exclude from our data, and who bears the consequences of our models. In this interactive session, David will share a practitioner's framework for ethical reasoning drawn from 20 years of building data-driven systems in complex, high-stakes environments, from cybersecurity operations protecting hospitals and NGOs to geospatial intelligence analysis and data systems that informed a national peace process. The session will explore how these experiences connect to the ethical challenges data scientists face today: AI governance and accountability, the concentration of power in the AI ecosystem, and the question of how to build technology that serves broad human flourishing. Rather than treating ethics as a constraint on technical work, this talk argues that ethical judgment is a dimension of professional craft, making data scientists more effective, more trusted, and more durable in their careers. Expect an interactive format with live polls, real-world scenarios, and open discussion throughout.

About the Speaker: David Schoeller-Diaz is a global program leader and systems builder with 20 years of experience leading data-driven operations across humanitarian, security, and technology sectors. His career spans intelligence analysis for U.S. Southern Command, humanitarian information management with iMMAP and the United Nations, cybersecurity partnerships at the CyberPeace Institute, academic research at the Harvard Humanitarian Initiative, and co-founding a social enterprise in Colombia. He holds a Master of Arts in Law and Diplomacy from The Fletcher School at Tufts University, an Executive Master in International Business from ESCP Business School, a PMP certification, an Executive Data Science Certificate from Johns Hopkins University, and a BA in Political Science from USF, where he was a McCarthy Fellow. David is based in the San Francisco Bay Area.

This talk is especially relevant for students interested in responsible artificial intelligence, trust and safety, policy, and governance.

We hope you can join us.
#DataScience #ResponsibleAI #AIEthics #AIGovernance #TrustAndSafety #USFCA #USFMSDSAI #DataInstitute #TechForGood #Cybersecurity #DataScienceSpeakerSeries