AGENDA Generative AI Applications Summit 2024

Phoenix, AZ June 4-7, 2024

In this keynote, you’ll learn how open-source models can help you build high-quality AI applications, generative or not, while giving you more flexibility, control, and ROI than closed-model APIs. We’ll highlight the latest and greatest models, and show you how to get started with them in minutes. Along the way, you’ll also learn about the technical ecosystem that Hugging Face is fostering, from models and datasets, to cloud integrations and hardware acceleration.

The greatest tool is the hardest to use. Machine learning is the world’s most important general-purpose technology – but it’s notoriously difficult to launch. Outside Big Tech and a handful of other leading companies, machine learning initiatives routinely fail to deploy, never realizing value. What’s missing? A specialized business practice suitable for wide adoption.

In this keynote, Machine Learning Week Founder Eric Siegel presents a six-step practice called bizML for ushering machine learning projects from conception to deployment. This disciplined approach, which is the topic of Siegel’s new book, The AI Playbook, serves both sides: It empowers business professionals and it establishes a sorely needed strategic framework for data professionals.

Generative AI (Gen AI) is transforming investment management, enhancing operational efficiency, personalizing client services, and fostering innovative asset management strategies. By utilizing language models and multimodal LLMs, firms can effectively extract valuable signals from news, filings, and various unstructured data sources, driving market share growth and streamlining research processes to make advanced analytics more accessible. As Gen AI can significantly improve risk management and enrich client interactions, its adoption has become essential for firms seeking competitive advantages in a dynamic financial environment. Drawing from my experience at the intersection of Gen AI and investment management, I will share the industry viewpoint and practical insights on harnessing Gen AI’s transformative potential.

Building successful, interactive generative AI applications for troubleshooting or assistive knowledge management system from complex data that is originally meant for human readability is not a trivial task. In an organization documents can range from trouble shooting technical manuals, knowledge articles on specific product descriptions with lot of tables, images, links, and some machine generated outputs like call transcripts. This presentation highlights the key lessons learnt in implementing complex infobots and trouble shooting bots in a multi-tired way to meet use case scenarios for enhancing productivity of technicians and customer facing agents.

Paychex manages a diverse range of textual data, including phone call transcripts, emails, chat logs, and survey feedback. A primary opportunity is to leverage this data effectively for actionable insights to improve the customer experience, and help small businesses succeed. The Paychex Data Science team employs a suite of large language models (LLM), covering traditional methods like word embeddings and phrase matching, to cutting-edge technology such as generative AI. Through these methods, we identify clients at retention risk, address economic distress signals, and categorize communications for product improvement. Our advanced LLM technology promises continuous enhancement to deliver tangible business value.

Discover the cutting-edge methodologies for preventing hallucinations in large language models (LLMs) used for chatbot development. This talk highlights the journey and key learnings from creating a custom chatbot, emphasizing the deployment of anti-hallucination strategies. Over the session you will explore the causes of LLM hallucinations, their impact on user trust, and effective strategies to mitigate them. This session offers a comprehensive toolkit for developers and researchers aiming to enhance the reliability of LLM applications, ensuring your projects maintain accuracy and credibility.

As a Generative AI Ambassador at Google, Hitesh brings a wealth of experience and knowledge to show how companies can scale their business, explore new innovative ideas, and drive value from GenAI. Having witnessed firsthand the incredible impact of Generative AI on businesses (Fortune 100), his talk will dive deep into the intriguing world of Generative AI, providing insights into its core concepts, capabilities, and cutting-edge developments. Besides, he talks about the GenAI value chain, and what is the potential opportunity for each one of us.

The Generative AI Accelerator It is a unique exercise that guides participants through leveraging Generative AI to drive business innovation. It is a proven framework, tested with over 25 organizations, ranging from small to medium-sized businesses to large corporations. In a two-step process, we first identify which problems are most suitable for solving with generative AI models. This is achieved by providing participants with prompting templates to test the feasibility of cases against AI models in real-time. We then draw shortlisted use cases on the GenAI Architecture canvas, accompanied by guiding questions, which aids participants in drafting their GenAI solution architecture. This process elucidates how a GenAI architecture can be optimally integrated into organizations, addressing the most pertinent use cases. Participants leave with a comprehensive understanding of how a GenAI solution would function within their specific organization, including which data to leverage and the essential elements of cloud services needed to build the solution.

Requirements: 

• Be aware of potential use cases in your area. 
• Bring a laptop. 
• Have an internet connection
• Have access to an intelligent chatbot e.g. Opus, ChatGPT, Gemini, or open-source variants.

Natural Language Generation has made substantial progress – you can easily generate most types of text without sacrificing creativity. But the more creative your prompts, the more likely you are to generate non-compliant language. The challenge is how to comfortably generate text in regulated industries like finance. Last discusses potential solutions.

In today’s digital age, the quality and accessibility of knowledge assets are paramount. Aramco recognizes this and has established a comprehensive corporate knowledge base that consolidates validated knowledge assets across various disciplines. But how do we ensure the quality of these assets and prevent the tedious cycle of reviewer rejections and resubmissions?

Amjad will unveil Aramco’s approach to maximize knowledge value unsing an in-house developed virtual assistant that not only provides users with guidelines and templates but also harnesses the power of generative AI to review the content and structure of knowledge assets, ensuring they meet a predefined criteria specific to each knowledge type. Moreover, it enhances the writing quality, creates abstracts, references, etc. making it more readable and reusable, thus amplifying its business value.

By participating in this session, attendees will:

• Gain insights into Aramco’s AI-Empowered Knowledge Base.
• Understand the role of AI in enhancing knowledge asset quality.
• Learn how to leverage AI capabilities to optimize content for maximum business impact.

Explore the latest in Generative AI, spanning text, image, video, and 3-D object generation, and uncover how these technologies are shaping the future of product development. We will examine the role of AI agents and their increasing capabilities. Additionally, we’ll investigate imminent developments edging us closer to artificial general intelligence (AGI), offering insights into achieving human-like intelligence in machines. This session will equip participants to understand current and forthcoming AI advancements and their impact on innovation.

Although LLM-powered applications are all the rage, the best practices to make them safe, reliable, and robust are still being determined. It is critical that high-value brands be able to experiment with these technologies without compromising their brand.

In this session, we’ll share a behind-the-scenes look at how Harvard Business Review launched a brand-aligned generative AI experience in their consumer and business-to-business markets. We’ll unveil the various concerns, approaches, and tradeoffs that our team of researchers and engineers explored, balancing principled research with pragmatic product development. We’ll leave attendees with a checklist of concerns to address for their own experiments.

Over the last year, generative AI has exploded out of the dark corners of the tech industry to become a central object of public attention, imagination and investment. As generative AI tools are increasingly integrated into widely used products, ML engineers shoulder significant responsibility for the safety, risks and social impact of the systems they build. On the other hand, data engineers are expected to integrate a variety of new AI based tools into development workflows and data pipelines. In this talk, I will address the longstanding cultural gap between the two disciplines, and share thoughts on how data and machine learning engineers can work together to meet the challenges and opportunities created by the AI boom.

Meet new people for dinner at a restaurant in walking distance and network in a relaxed setting with good food and favorite beverages (everyone pays for their own dinner). We will publish the restaurants closer to the date.

For those of us who’ve spent any time arguing with customer service chatbots or screaming “REPRESENTATIVE!” to an interactive voice response (IVR) line, the hype around generative AI improving customer experience might seem hard to believe. Sure, ChatGPT can do some impressive things, but do we really think AI can make the painful process of dealing with our insurance companies, wireless carriers or utilities any easier? The answer is yes, but not by building a better bot.

This presentation will explore how the real breakthrough created by GenAI is making the entire operation smarter and more personalized, so there are fewer reasons to deal with customer service in the first place.

As a Gen AI leader in the banking industry, Maria is excited to share insights on how Generative AI is transforming the way banks operate and engage with their customers.

In this session, we will explore the immense potential of Gen AI in banking, focusing on its ability to drive tangible business value. We will delve into real-world examples of how Gen AI is revolutionizing customer engagement, enabling personalized experiences, and enhancing operational efficiency.

Through interactive discussions and case studies, we will examine the key challenges and opportunities that arise when implementing Gen AI in banking. We will also explore the ethical considerations and regulatory frameworks that need to be addressed to ensure responsible and secure use of Gen AI technologies.

Join Maria in this session to gain a comprehensive understanding of how Gen AI is reshaping the banking landscape, and discover strategies to leverage its power for your organization’s success.

This talk summarizes current trends and use cases that apply state-of-the-art natural language processing in healthcare systems, pharmaceuticals, and health IT companies. It covers unlocking use cases that become possible with current large language models, enabling medical domain experts to train & tune models in a co-code environment, and testing NLP models across aspects of Responsible AI in preparation for regulatory oversight. In this talk, David also shares free and open-source tools you can use today in these three areas.

Most businesses have several Million Dollars invested in physical assets which range from transformers, batteries in utilities, manufacturing equipment, fleets, buildings and others. Constant maintenance and upkeep is paramount to the operations team. Generative AI transforms we have done asset management, allows leveraging not only large language models but Time series Foundation Model and Computer Vision Foundation Models for rapid adoption of AI in cases where we have less data, makes it more manageable and applied with less effort. Use Cases like Field Assistance, Advisor for Asset Management, Failure Mode Understanding and Strategy Generation, Anomaly using Generative AI & Few-Show Vision based anomaly detection.

In the chat-bot and generative AI industry, there is strong demand for Text-to-SQL tool that can take user input questions, create SQL queries, and return results of those queries back to the user. Successfully creating a training dataset then comparing and choosing the right LLM to put into production can be difficult. This session demonstrates how to successfully curate a Text-to-SQL training dataset and compare, evaluate, and fine-tune various Generative AI LLMs to ensure the highest accuracy and customer satisfaction.

Neural Networks are flexible, powerful, and often very competitive in accuracy.  They are criticized primarily for being uninterpretable black boxes, but their chief weakness is that backpropagation makes them unrepeatable.  That is, their final coefficient values will differ, from one run to the next, even if the NN structure, meta-parameters, and data are held constant!  And unlike multi-colinear regressions, the varied NN coefficient sets aren’t just alternative ways — in an over-parameterized model — of producing similar predictions.  Instead, the predictions can vary a disquieting amount and often “converge” to a significantly worse training fit than is possible.  

What happens if one instead employs a global optimization algorithm to train a NN?  Untapped descriptive power should be unleashed, encouraging use of simpler structures to avoid overfit.  And, with randomness removed, the results will be repeatable.  We’ll demonstrate initial results for (the relatively small) NNs practical to optimize.

Join Paul in this engaging talk as we demystify Gen AI and uncover strategic approaches for traditional retailers to navigate the ever-changing retail landscape. While Gen AI offers incredible opportunities, it comes with its set of challenges and concerns that we’ll address together. We’ll delve into the complexities of integrating AI in retail, discussing issues like data privacy, ethical considerations, and the imperative to upskill our workforce. Together, we’ll explore the vast opportunities Gen AI presents, from crafting hyper-personalized experiences to leveraging data-driven insights that can redefine customer interactions. This talk is not just about strategies; it’s a journey into understanding and harnessing the power of Gen AI, embracing its potential, addressing concerns, and ensuring traditional retailers thrive in this transformative era.

Explore the intricate dynamics between Generative AI (GenAI), predictive analytics, and prescriptive modeling. Witness how this synergistic triad reshapes our understanding of AI’s capabilities, moving beyond perceived magical realms. The session delves into real-world applications, demonstrating the transformative impact of integrating GenAI with predictive and prescriptive approaches. Brace yourself for an exploration of AI’s true potential, demystifying illusions and ushering in a new era of possibilities.

This presentation discusses a case study for leveraging AI-based knowledge engagement techniques for development and use of Immunization Information Systems (IIS) guidelines. The knowledge base developed by the IIS community is voluminous and growing rapidly, resulting in a valuable yet difficult to navigate resource. Initial results of employing a generative AI tool (ChatGPT) at the CDC’s National Center for Immunization and Respiratory Diseases (NCIRD) are promising and indicate increased accessibility and discoverability of information. By making all guidance documents searchable at once, discoverability and presentation of content are enhanced. The AI tool uses a limited set of trusted in-house sources: guidance manuals, community standards, emails, and meeting notes. This ensures accuracy of responses and guarantees information is only from verified sources. Additional capabilities include support for new content development with information synthesis across multiple resources.