Talk - Systems with Intent: Designing Agentic AI for Financial Decisioning

#computer #society #silicon #valley #bigdata #datastreams #WIE

Hosted via Google Meet -  https://meet.google.com/pqw-trcy-xpg

The financial industry is entering a new phase of intelligence—one where AI systems are no longer just analytical tools but autonomous collaborators capable of interpreting goals, reasoning over complex data, and taking context-aware actions. This talk explores the emergence of agentic AI systems in finance—AI architectures designed with intent, memory, and decision agency at their core.

We’ll examine how these systems differ from traditional automation by introducing the concept of goal-driven autonomy: agents that plan, execute, and self-reflect across dynamic financial environments. Drawing from real-world applications such as loan underwriting, portfolio monitoring, and compliance intelligence, the session will unpack how planners, executors, and reflectors interact within an agentic loop to make transparent, auditable decisions.

The discussion will also highlight enabling technologies—large language models as reasoning engines, retrieval-augmented generation (RAG+) for contextual grounding, and knowledge graphs for structured memory—and how they converge to form adaptive decision frameworks. Finally, we’ll address the critical dimensions of safety, alignment, and regulatory oversight required to operationalize agentic AI responsibly in financial ecosystems.

Participants will gain a systems-level understanding of how to engineer intent into AI, design autonomous yet trustworthy financial agents, and prepare for a future where decision-making is increasingly shared between humans and intelligent systems.

By registering for this event, you agree that IEEE and the organizers are not liable to you for any loss, damage, injury, or any incidental, indirect, special, consequential, or economic loss or damage (including loss of opportunity, exemplary or punitive damages). The event will be recorded and will be made available for public viewing.