Dropping Datadog Without Losing Your Bark: Open-Source Observability

Observability is supposed to help us understand our systems but for many teams it has quietly become one of the most expensive and least flexible parts of the stack. Plenty of organizations take polished SaaS platforms like Datadog to get real value quickly early on, only to hit the same wall later: runaway costs, pricing tied to architectural decisions, and vendor lock-in so strong it makes Rogers and Bell both blush.

In this talk, I’ll walk through a real-world case study of a JVM-based production system that migrated away from a Datadog-centric observability setup to an open-source, OpenTelemetry-based platform. We’ll dig into the why behind the move, the architectural decisions that mattered, the mistakes along the way, and how the team ended up with pragmatic visibility into latency, errors, and resource usage at a fraction of the cost.

The second half of the talk is practical and forward-looking. I’ll show how OpenTelemetry unlocks real choice by acting as a neutral data plane for your telemetry. From there, we’ll look at modern, off-the-shelf observability backends you can plug in within minutes without rewriting applications or installing new agents. Want to dual-write, swap vendors, or route data somewhere new? This is how you do it without a big-bang rewrite or another long-term commitment with a new vendor.

This session is for engineers and CEOs who are feeling the observability cost squeeze, want leverage back from vendors, and are looking for realistic ways to modernize monitoring without lighting their entire stack on fire each time they decide to change their tools of choice.

About the venue
Free Times Cafe has bistro-style seating and a full food and drink menu. Please consider helping to support the venue by planning to have supper during the talk.

Speaker Bio
Alexei Zenin is a Platform Engineering Consultant based out of Toronto, Canada. Over the course of his career he has worked across the Healthcare, Marketing, Video Games, and Banking industries. He loves to tackle hard problems in the distributed systems space with a focus on scaling platforms both from a technical and process perspective. He is a strong advocate for highly maintainable and well tested code while at the same time pragmatically delivering value to the end user.

He has contributed to the broader tech community, namely to Confluent’s blogs on Apache Kafka, spoken at Current 2022 in Austin, Texas, and spoken at XP Game Summit 2025 in Toronto, Canada. During his 5 years at Uken Games he contributed to several games like Who Wants to Be a Millionaire and Jeopardy. One title of note he saw from inception to completion was a game for Tim Hortons built right into the official app, Tim’s Word Challenge. Millions of Canadians have played it along with their morning coffee. He also got to work with Netflix on several unannounced games during his time at the studio.

Currently, Alexei is working with various clients, both large and small to modernize and streamline their platforms with the most appropriate tools and practices. Outside of his professional life, he gives back to his community by volunteering at UofT. He also enjoys paddleboarding and attending live rock concerts to reset after long coding sessions.