Skip to content

Talk Proposal: What Makes Agents Different? #20

@usize

Description

@usize

Talk Information

  • Draft Title: What Makes Agents Different?
  • Length:
    • [ x] Lighting talk/Demo (10 min) (I can happily condense the talk if necessary)
    • [ x] Full Presentation (20 min)
  • Links: Short blog post covering the same topic
  • Which dates, particularly Tue/Wed/Thu during the 3rd/4th week of a month, are you likely available? I don't have any standing dates that won't work after December 2025. Otherwise, I'm busy the week of December 15th.
  • Short Summary of your talk:

In this presentation, I argue that a "software agent" -- especially one built on an LLM -- isn’t just another backend service; it’s fundamentally a different kind of system, defined by runtime identity, stochastic behavior, and context-dependent decision loops. I’ll show why an agent is properly viewed as a gestalt of components (model, tool-set, memory/context, goal) rather than just glue code, and how that distinction matters for deployment, auditing, and reliability.

Speaker Bio

  • Name: Morgan Foster
  • Mini-bio:

Morgan works in Emerging Technologies for Red Hat, where she focuses on developing areas of active research into usable products. She is a maintainer of the Kagenti project, and participates in the AI Gateway Kubernetes workgroup, extending the Kubernetes Gateway API to better support AI use cases (e.g. Guardrails).

Metadata

Metadata

Assignees

Labels

Type

No type

Projects

No projects

Milestone

No milestone

Relationships

None yet

Development

No branches or pull requests

Issue actions