Breakout Sessions

How do you enable a team of 100 engineers to effectively support 300+ critical applications across five hosting platforms? At Thomson Reuters, we turned to AI - not as a buzzword, but as a genuine force multiplier. Experience our journey of transforming traditional IT operations into an AI-augmented powerhouse, where Datadog, ServiceNow, and custom AI solutions work in harmony to create a next-generation control plane. We'll share real victories, honest challenges, and practical insights from our mission to build a more intelligent operational framework.

As platforms scale, so do the challenges of monitoring and incident management. How do you ensure your monitoring strategy evolves with growth? At Toyota Connected, we launched our Telematics Service Platform six years ago. Since then, we've deployed new services, expanded across global regions, and welcomed new team members. Utilizing Toyota's Kaizen philosophy (continuous improvement), we streamlined our monitoring process to keep simplicity at scale. In this session, I'll share the key strategies we've implemented to ensure service reliability. This includes bi-weekly monitor review meetings, resiliency and performance testing through gamedays, and streamlined incident management workflows. Learn how these continuous improvement practices help us catch issues early, reduce mean time to resolution, and empower teams to build with confidence, all while scaling effectively.

It seems that everyone has an opinion on AI—sometimes wildly enthusiastic or pessimistic—but often with very little grounding. So how can you filter out the speculation from reality? For this session, we’ve brought together two deeply knowledgeable experts, Ameet Talwalkar (Datadog Chief Scientist, CMU Professor) and Zico Kolter (CMU Machine Learning Department Head, OpenAI Board Member) to talk about the current state and future trajectory of AI. They’ll discuss where AI is excelling, where it’s falling short, and what we can look forward to in the future based on current research. You’ll leave with a clearer understanding of how to apply AI to your work today and how to prepare for its advancement in the future.

Kubernetes cluster administrators are frequently faced with the complaint "My service is broken." But what is broken? Is it the service or the underlying platform? How long will it take to find the root cause? Why wasn’t the platform issue caught before it affected services? In this talk, Jon Winton will discuss how platform administrators can automate user journey testing within Kubernetes platforms, powering Service Level Objectives to identify regression in the system before they affect users. Whether running one or hundreds of Kubernetes clusters, measuring performance with automated testing can lead to more efficient and less buggy platform iteration. Attendees will learn how to execute these tests and surface telemetry around user journeys, allowing platform administrators to monitor the "golden paths" of their platforms.

At Coinbase, maintaining service reliability at scale requires navigating complex service dependencies, pinpointing root causes, and minimizing incident mitigation time—challenges amplified by massive volumes of telemetry data. In this presentation, we’ll explore how Coinbase uses distributed tracing alongside graph analytics in Databricks to build a service dependency graph, and how better understanding service relationships has improved reliability. Designed for SREs, platform engineers, and observability teams, this session will provide insights into enhancing the root cause analysis (RCA) processes and making better use of Datadog telemetry data.

As AI agents rapidly improve, becoming more autonomous and interconnected, they unlock new ways to assist us. But as they perform actions for us and delegate tasks to other AI agents, we need to reexamine our understanding of “identity.” How do we ensure these powerful AI interactions are authentic, authorized, and permissioned, while differentiating between legitimate actions and potential misuse? Join Datadog co-founder and CTO Alexis Lê-Quôc and Okta CTO Bhawna Singh as they explore the convergence of AI, security, and observability. They’ll delve into the complexities of agentic AI and where they see the risks and opportunities. They’ll also share how they’re leading their teams as they harness this burgeoning technology to accelerate engineering performance. You’ll walk away with actionable advice to unify observability and security, infusing trust into your AI-augmented systems and maximizing your organization’s overall impact.

Today's security leaders navigate the challenge of fostering innovation and speed while safeguarding system and user trust. In this fireside chat, Datadog CISO Emilio Escobar and a special guest will discuss defining and calibrating acceptable risk levels within their organizations. They will explore practical decision-making that allows teams to accelerate delivery without sacrificing fundamental principles of trust, security, and compliance. This candid conversation will cover real-world risk trade-offs and guiding frameworks, offering insights into leading secure, high-velocity organizations.

There are many factors that contribute to creating a high-performing engineering team, such as processes, automation, and culture. While all of these are important, the most critical ingredient is engineers. But how do you recruit and hire the best engineers? Join Datadog’s Ambra Benjamin (VP of Tech Recruiting) and Jeromy Carriere (SVP of Engineering) as they chat about their approaches. They’ll share advice for recruiting, interviewing, retaining, and mentoring to help you build exceptional engineering teams.

According to some studies, nearly 90% of companies are running hybrid or multi-cloud. As your IT footprint grows, so does its complexity and the likelihood for hidden gaps in security. In this session, Datadog’s Wissal Lahjouji will chat with Robert Wise (Cognizant Trizetto), Paul Richards (Citizens Bank), and Bobby Argeanton (Hyatt) about how they protect critical data across legacy and modern environments while meeting compliance demands. They’ll share ways to assess costs, skills, and resources in multi-cloud security. You’ll learn how proper planning keeps operations smooth and avoids disruption, and leave with structured methods to maintain trust and reduce risk.

Transitioning from a legacy SIEM provider to a new SIEM comes with many challenges, but for MyFitnessPal it was worth the effort. After struggling with administrative challenges, siloed tooling, and noisy false alarms, MyFitnessPal decided to switch to Datadog Cloud SIEM. In this session, Allen Cox will share the journey of the migration and discuss how Datadog addressed their previous pain points. He'll also discuss some of the unexpected benefits of using Cloud SIEM and the impact it's made across MyFitnessPal's engineering organization. You'll learn how Datadog's Cloud SIEM can optimize your security operations and Allen's tips for achieving a seamless migration experience.

Have you ever found yourself operating a system you aren’t familiar with? Maybe it was written in a really old framework. Maybe it was a language you don’t understand that came through an acquisition. Did you find yourself getting frustrated with slow delivery, integration test overload, duplicate entities, and architectural dead ends? You try everything to understand and improve these old systems, but eventually you realise it’s fruitless. You need to modernise! But where do you even start? This is the situation we found ourselves in at Flipdish in 2023. We were operating with not one, but two monoliths - our original .NET (Framework!) system and a second JavaScript stack inherited through acquisition. This talk shares how we scaled modernisation from a single serverless pilot to a company-wide platform approach. We adopted AWS serverless, embraced event-driven architecture, and built a Golden Path with reusable templates, governance-as-code, and shared tooling to reduce friction. Datadog acted as a key platform partner throughout, giving us visibility across monoliths and our new Domain Services, tracking CI pipelines and cloud costs, and enabling a consistent view across fragmented systems. It helped align teams and accelerate change without sacrificing control. You’ll hear what worked, what didn’t, and what we would approach differently if we started again. This isn’t a polished success story - it’s a practical account of how we’re moving from tightly coupled monoliths to independently owned, event-driven domain services. If you’re modernising a complex system or coordinating change across multiple teams, you’ll leave with clear, tested strategies for maintaining momentum, ensuring visibility, and enabling long-term architectural evolution - without disrupting day-to-day delivery.

What happens when your platform faces 30 times more traffic than usual — and the regular dashboards aren't enough? Twice a year, our e-commerce platform enters a high-stakes period we call "the Archives" — an intense sales event that pushes our systems to their limits. During our most recent edition, we broke all previous records… and hit performance ceilings we had never encountered before. In this talk, I'll share how our team prepared for the unexpected, how Datadog helped us cut through the noise in real time, and how we detected a subtle yet critical issue that could have gone unnoticed — hidden in a storm of data. You'll walk away with concrete ideas on how to monitor for edge cases, design alerts that matter during chaotic traffic spikes, and build resilience for days when "normal" doesn't apply.