Operations | Monitoring | ITSM | DevOps | Cloud

Multi-Agent Architectures - What we shipped, what broke, and what we'd do differently

At LLMday Lisbon, our Software Engineer, Viktor Vasylkovskyi, highlights the realities of building production AI agents with LangGraph - sometimes getting it right, often learning the hard way. This talk is about what was actually shipped, including a distributed multi-agent setup at PagerDuty. Viktor breaks down the real tradeoffs between LLM-driven and deterministic orchestration, what broke, and how he’d approach it differently now.
Featured Post

From firefighting to forward planning: a practical route to operational innovation

Operational innovation is often treated as a back-office efficiency exercise, but in practice, it is becoming a strategic discipline. As AI moves deeper into day-to-day operations, technical leaders need a clearer way to cut toil, reduce risk and build the capacity to innovate. For many operations teams, it starts with incident management. When responders are trapped in noisy alert streams, manual escalations and fragmented workflows, innovation is pushed aside by the urgent work of keeping services available.

Resilience for an AI-Powered Future: PagerDuty's FY26 Impact Report

The impact vision for PagerDuty.org is to enable mission-driven teams to build a resilient world and a sustainable future for all. As a leader in modern, AI-first operations, we know that operational excellence supercharges social impact. As artificial intelligence rapidly reshapes the social sector, this commitment to resilience and efficiency has never been more vital.

What Major Incidents Really Cost Your Business

When a major IT incident hits, most organizations know what it costs in the moment: lost transactions and missed SLAs. But according to the findings of our 2026 State of AI-First Operations report, the most significant consequences often don’t show up until long after the incident is closed—in customer relationships, team health, and brand reputation.

PagerDuty Report Finds Two-Thirds (66%) of Office Professionals Have Used Unauthorized AI Tools at Work

Three-quarters of office professionals (75%) say they would be likely to look for a new job that offered better AI skills development, a figure that climbs to 80% at companies with $1 billion or more in revenue.

Shadow AI Is Happening Within Your Organization

A majority of office professionals (72%) believe they understand how to use AI for their job better than the team responsible for managing AI at their company. While it’s encouraging to see employees embrace AI with such confidence, organizations will want to ensure they are providing the tools, guidance, and safeguards needed to help employees use AI safely.

Behind the Scenes: Shift-Based Schedules

The PagerDuty team lifts the hood on the newly rolled out Shift-Based Schedules. This session breaks down how PagerDuty is moving away from layer-based architecture to a flexible system that natively scales with modern engineering teams and naturally fits their workflows. Timestamps: Speakers: Ken Choate (Software Engineer) Kelsey Yocum (Sr. Product Designer) MJ (Sr. Engineering Manager) Todd Murphy (Principal Product Manager)

Insights Agent: Deep operational intelligence where your team works

This blog post is part of PagerDuty’s ongoing series on how we’re helping customers navigate their journey towards autonomous operations. Read on to learn about how PagerDuty Advance Insights Agent (now Generally Available for Microsoft Teams users) builds towards this vision. As AI accelerates development and teams ship more code than ever, operational data is everywhere; insights aren’t.

Scribe Agent updates: no more manual note-taking or lost context

This blog post is part of PagerDuty’s ongoing series on how we’re helping customers navigate their journey towards autonomous operations. Read on to learn about how PagerDuty Advance Scribe Agent updates (Generally Available) build towards this vision. When a major operational issue hits, there’s always someone drawing the short straw to take on the most thankless job in incident response: scribing the call. Chances are you were already that someone.