In the era of generative AI, traditional machine learning (ML) may seem overshadowed, yet it remains a cornerstone of modern AI. This talk explores the enduring relevance of ML in a world increasingly dominated by generative models. While generative AI excels in creative and content generation tasks, it relies heavily on foundational ML techniques for data preprocessing, feature extraction, and model evaluation. The talk will delve into how ML continues to play a critical role in areas like predictive analytics, anomaly detection, optimization, and decision-making—domains where generative AI is less effective. By examining the symbiotic relationship between ML and generative AI, this presentation will argue that far from being obsolete, traditional ML methods are essential for enhancing and guiding the capabilities of generative AI, ensuring robustness, interpretability, and real-world applicability.

Abstract coming soon.

Enterprises hold a wealth of data, much of which remains untapped. In today’s data-driven landscape, unstructured data—spanning text, images, videos, and more—comprises over 80% of all enterprise data. Yet, its full potential often goes unrealized. In this session, I will explore strategies to unlock the value of unstructured data, enabling actionable insights, enhanced decision-making, and innovation. Drawing from a real-world use case, I’ll demonstrate how raw audio recordings can be transformed into a structured format to drive positive business outcomes. I will also address the key challenges encountered during this process and share practical tips for overcoming them effectively.

This talk encapsulates the essence of exploring diverse time series forecasting methods across sectors such as travel, logistics, supply chain, and retail. It highlights the comparison between traditional econometric techniques and modern machine learning approaches like deep learning, transformer models, and LLM-based methods. The talk would delve into how these varied approaches are applied in different industries to enhance forecasting accuracy and decision-making processes.

Abstract coming soon.

To improve street conditions and address the homelessness crisis, the City of San Francisco employs over a dozen specialized street outreach teams. These teams reverse overdoses; provide mental, physical, and behavioral medical care; make housing referrals; and, more broadly, improve the street conditions and public safety of the city. These efforts are often heroic, yet there are still people on the street in crisis.

The promise of data science is that we have tools to use data to improve these and other civic challenges: we have algorithms to fuzzy match across different departments’ databases; we have models to predict outcomes and intervene with high-risk individuals; we have analyses and trials to learn what methods work best.

How does this work in practice, though? What are the legal, organizational, procedural, and technical challenges in actually levering this data? In this talk, Dr. Amanda Ford explains how her team tackled these and other issues, and improved homeless outreach in San Francisco.

Abstract coming soon.

Generative AI is revolutionizing industries, but turning its potential into tangible results requires more than just ambition—it demands a strategic approach. This session explores how to identify and prioritize the most impactful GenAI opportunities within your organization, ensuring your efforts are focused on initiatives that truly drive business success. Attendees will gain insights into crafting a clear roadmap to guide their GenAI journey and inspire confidence among stakeholders.