CEO Fireside at HumanX: Resilience at the Speed of Change
PagerDuty CEO and Chairperson Jennifer Tejada in conversation on April 8, 2026 at HumanX in San Francisco with Honeycomb CEO Christine Yen and journalist Jennifer Strong, show how observability and real-time response help builders spot issues sooner, fix them faster, and learn from every incident.
Chapters & Topics:
0:17 Session opening and context
The moderator introduced the session and framed the topic as reliability in fast-changing distributed systems, welcoming CEOs from PagerDuty and Honeycomb to discuss detection and response at scale.
1:25 Defining reliability and modern failure modes
Panelists defined reliability as a boardroom and product-level concern, distinguished prevention from resilience, and described modern failures as novel, partially understood, or well-understood while noting predictability limits.
- Reliability and resilience are separate but interdependent priorities that now require CEO- and board-level attention.
- Modern failures are shifting from downtime to degradation and disrupted brand experiences.
3:49 Speed, change, and uncertainty
Participants debated whether speed or the nature of change is harder to manage, agreed both matter, and emphasized that uncertainty and non-determinism require new habits and practices in software development.
- Change velocity and the novelty of failures both drive the core tension in achieving reliability.
6:40 Trust, transparency, and customer impact
The group stressed that trust is earned through consistent transparency, good debriefs, and communication about impact and remediation, and that brand perception suffers regardless of root cause.
- Trust is built through transparency, blameless postmortems, and clear post-incident communication.
10:40 Incident teams and culture
Discussion focused on what makes incident teams effective, highlighting psychological safety, collaboration between humans and agents, and using incidents as leadership and learning opportunities.
- High-performing incident teams treat incidents as learning opportunities and practice via tabletop exercises.
17:38 Postmortems and operationalizing learning
Panelists recommended data-driven postmortems with human context, follow-up practice such as tabletop exercises, and warned against superficial fixes that address symptoms rather than root causes.
- AI adds a new risk layer for enterprises and requires tailored governance for regulated organizations.
- Automated agents can handle well-understood problems, but human context remains essential for novel incidents.
24:53 AI, automation, and leadership trade-offs
The conversation closed on AI as both a productivity opportunity and a new risk layer, underscoring the need for governance, careful adoption across regulated and unregulated environments, and continued human oversight.
Subscribe to our channel: https://bit.ly/3BNQYNS
Follow us on Social:
Instagram: https://bit.ly/3xf96g8
Facebook: https://bit.ly/3zCUM2g
Twitter: https://bit.ly/3f4PC7p
LinkedIn: https://bit.ly/3f3Ex6R