AI is a rapidly growing field that has the potential to revolutionize many industries. However, many people are still unsure of what AI is and how it can be applied to their environment. Seth Earley shares his knowledge and expertise gained over 25 years of consulting about how new and emerging technologies can be safely and effectively deployed in the enterprise. Large Language Models and ChatGPT types of applications are getting a great deal of attention, but how can organizations use them to retrieve information across the entire enterprise, speeding information flows and creating new efficiencies and capabilities? AI solutions require having the correct data and the correct information architecture. Seth explains how this can be achieved at scale saving hundreds of millions of dollars in costs and increasing revenue by the same order of magnitude through case studies and examples. For a pharmaceutical company, the use of a knowledge architecture improved accuracy from 53% to 83% and eliminated hallucinations for a highly sensitive portfolio review application while providing audit trails, protection of IP and explainability. Seth educates and level sets with a warm, humorous, and engaging style accessible to both executives and practitioners.
Digital transformation has become a broad, catch all term meaning anything from improving the customer experience to streamlining internal business processes and everything in between. Artificial Intelligence has also become a broad and ambiguous concept meaning everything from chatbots to self-driving cars. Now put those two concepts and you end up with jargon soup, wasted resources, confusion, and failed initiatives. The industry is full of hype and confusion, buzzwords and bs, unrealistic expectations and some high visibility failures.
Whenever a significant technology change comes about, it’s difficult to not get caught up in the confusion and end up on costly paths without business value due to the fear of missing the next big thing and losing out to the competition
In this session we will hear from award winning author and popular conference speaker Seth Earley who will provide practical guidelines for getting real business value from Artificial Intelligence programs and projects and discuss what needs to be in place to be successful (hint: it’s not the technology).
Attendees will leave with a buzzword free understanding of:
- The necessary business decisions when considering these tools
- Cultural issues and how to address them
- How to approach processes to maximize business value
“We need more training data” is a common reframe from AI project principles. However the organization does not lack data. If anything there is too much data and machine learning and automated approaches are needed to understand and make use of enterprise data. The missing ingredient is a consistent framework and structure for information. Frequently the conventional answer is “master data”. While master data is important and critical to many processes, starting at the data level can cause the organization to miss important elements. A better place to begin is defining the concepts and things that are important to the business. Those concepts can then be designed into multiple systems and technologies.
This is a critical distinction that business leaders need to understand rather than leave to technology groups or worse, try to outsource to third party and off-shore providers.
One expression of these business concepts is referred to as an “ontology” which is the master knowledge scaffolding for the enterprise. Building, applying and maintaining an ontology sounds complex but it is no more complex than other important business considerations.
Creating an ontology is an essential investment to prepare your enterprise to realize the benefits of AI and machine learning. Gone are the days when businesses should simply allow a number of small AI projects to blossom independently: for these projects to be competitive they need to draw on data from across the company, data stored in many different forms in many different systems and in different structures. An ontology defines these connections in a way that AI can take advantage of.
Business will be best positioned to build and apply practical business-focused ontologies if they 1. Identity pain points and friction in the current process 2. Define the desired future state 2. Understand what data is required to support that future state 3 Develop the organizing structures to harmonize and apply are most needed-before beginning to set the organizing principles for the ontology itself.
Attendees will leave with an understanding of:
- The business value of an ontology
- A non-technical explanation of how to develop an ontology
- How to structure data governance
- How to apply and gain efficiencies from this approach
With all the noise and hype around generative AI and ChatGPT, a crucial part of the puzzle is being glossed over: AI runs on knowledge – expertise and content structured and organized for retrieval using virtual assistants, chat bots, and conversational commerce applications. Knowledge and component content is also at the core of personalization algorithms and approaches. The ecommerce journey is a knowledge journey – at each step of the process users need answers to questions. What are your products? Which ones are appropriate for me? How do I make a selection? How do I install, use, support, etc.?
Many organizations are looking to ChatGPT and LLMs to provide such capabilities, but these tools still require foundational knowledge and content as well as correctly structured, high quality product information.
But how do organizations build these capabilities cost e!ectively and sustainably? AI and ML can help, but processes have to be optimized and the organization needs the correct reference architecture foundation upon which to build. The answer is a “knowledge factory,” an approach that solves information problems today while providing a foundation for the future. An approach for analyzing business problems, breaking problems into specific information requirements, developing and structuring knowledge components, and building out models for continuous learning and knowledge curation using automated and human in the loop approaches along with process metrics to quantify business impact. Organizations that do not build these capabilities will be caught flat footed as the industry progresses and competitors build sophisticated capabilities that enable higher levels of customer service at lower costs.
In this session we will walk through examples of how organizations have developed and applied a knowledge architecture for high functionality chat bots and virtual assistants as well as for personalized employee and customer experience.
• Leveraging data to electively segment your customer base and provide personalized suggestions and product offerings
• Driving personalization from promotions to pricing
• Predicting future uses for AI to enhance your customer relationships
Cognitive AI runs on knowledge – expertise and content structured and organized for retrieval using virtual assistants, chat bots, and conversational commerce applications. But how do organizations build this capability cost effectively and sustainably? The answer is a “knowledge factory,” an approach that solves information problems today while providing a foundation for the future methodology for analyzing business problems, breaking problems into specific information requirements, developing and structuring knowledge components, and building out models for continuous learning and knowledge curation using automated and human in the loop approaches along with process metrics to quantify business impact. Organizations that do not build these capabilities will be caught flat footed as the industry progresses and competitors build sophisticated capabilities that enable higher levels of customer service at lower costs.
Attendees leave with an understanding of:
- Actionable steps to begin their journey.
- How to develop a realistic achievable, roadmap based on current capability maturity
- How to apply a repeatable, scalable “factory” process to conventional knowledge management problems