OOPS #16 - Humans in the Loop: The Future of Incident Response
Details
A 'Humans in the Loop' community event with Elastic, Uptime Labs & PagerDuty in collaboration with OOPS!
We’re taking over the Greenwich Design District for an evening dedicated to the unsung heroes of the stack: The Humans in the Loop. AI adoption offers a range of benefits, but while our systems get more complex, the role of the human in incident response has never been more critical.
Come trade your real-time operations stories, share your best 'oops' moments, and learn how to keep the 'human' in your high-availability architecture.
📅 Date and Time:
Monday, June 1, from 6:00-9:00 PM
📍 Location:
Design District
Address: 19 Soames Walk, London SE10 0AX, UK
What to expect:
- 🎤 Lightning Keynotes - short, sharp and stimulating, local experts share their takes.
- 🍻 Sip & Savour: Delicious pizza and refreshing drinks to fuel the conversation.
- 🤝 The OOPs Group: Connect with the outage ops + incident response community, because misery loves company, but resilience loves a party.
🧠 Talks
Many Humans in Many Loops by Carly Richmond, Developer Advocate Lead (Elastic)
Human in the Loop, or HITL, provides us the ability to validate AI output using friendly and knowledgeable humans. But is that the only human loop in AI incident management systems? Join Carly as she covers other human loops, including reinforcement learning with feedback and context engineering, and how loops with a low threshold encourage alert fatigue.
Incident Response Reimagined: Accelerating Resolution with AI Agents by Daniel Afonso, Senior Developer Advocate (PagerDuty)
Learn how PagerDuty is leveraging Agentic AI to transform the incident lifecycle from reactive firefighting to proactive prevention.
Daniel demonstrates how new tools like the SRE Agent and Scribe Agent assist engineers during high-pressure outages by autonomously triaging alerts, querying logs in tools like Grafana, and transcribing context directly into incident channels.
Beyond immediate response, this demo highlights a "shift left" strategy using the new PagerDuty MCP server and Claude Code plugin, which allows developers to assess the risk of code changes against historical incident data before deployment, ultimately aiming to reduce the frequency of on-call.
✍️ RSVP now before we reach capacity
