As artificial intelligence continues to advance, there is growing interest in creating AI agents that can serve as subject matter experts across diverse domains. This talk will explore the challenges and approaches involved in building an AI system with deep knowledge and reasoning capabilities in a specialized field.

We will discuss techniques for curating and ingesting large corpuses of domain-specific data to seed the agent's knowledge base. Approaches to representing and organizing this structured and unstructured information will be covered. The talk will also delve into methods for enabling question-answering and generating substantive outputs leveraging the ingested domain knowledge.

Key challenges around knowledge updating, guarding against biases, and responsible development of these potent AI assistants will be explored. Real-world use case of implementation of such technology at UKG will be explored.

The talk aims to provide a comprehensive overview for practitioners interested in pushing the boundaries of artificial general intelligence towards usable domain-specific expertise. A Q&A session will allow for interactive discussion on this important emerging area of AI development.