About this event
As AI agents become a new class of users, traditional interface design principles matter more than ever. In this session, we’ll explore why building MCP-powered experiences isn’t just about connecting tools. It’s about creating interfaces that agents can effectively understand, navigate, and reason through.
We’ll discuss how to measure agent performance, manage context, think through security and access boundaries, and apply lessons from decades of interface design to this new era of AI. Through real-world examples and practical insights, you’ll walk away with a framework for building AI agent experiences that are more reliable, effective, and secure.
Attendees will learn:
- Why AI agents should be treated as users when designing MCP experiences
- How to measure and improve agent performance
- Best practices for context management and permission design
- Lessons from past technology shifts that apply to AI agents today
Hosted by
-
Team member
Kaleb Pomeroy Staff Software Engineer @ Specter OpsT
Kaleb is a software engineer at SpecterOps researching agentic systems, memory architectures, and interaction design. His work focuses on bridging modern AI research with practical software engineering.
-
Team member
John Hopper VP Engineering @ SpecterOpsT
John Hopper is a software engineer with two decades of experience spanning low-level Linux, cloud-scale services, and complex data platforms. His work focuses on reducing hard systems to clear abstractions and rebuilding them as practical, reliable tools. Most recently, he has helped shape the graph analysis and data foundations behind modern identity security through BloodHound and DAWGS. He brings a systems perspective to infrastructure, distributed data, and operational scale, emphasizing what technology enables people to understand and accomplish. He maintains open-source projects at https://github.com/zinic.
SpecterOps
Know your adversarySpecterOps provides adversary-focused cybersecurity solutions and training, enabling organizations to understand and defend against the tactics of threat actors through tradecraft analysis and attack path management.
