In the high-stakes world of enterprise integration, downtime isn't just an inconvenience, it's a breach of trust. For Digibee, an Integration Platform as a Service (iPaaS) connecting critical software systems, the operational reality is intense. They run over 18,000 pods, processing transactions where accuracy and speed are non-negotiable.
Tiago Bernardinelli, Engineering Director at Digibee, describes their daily reality with a blunt truth that every DevOps engineer understands:
Daily, our environment is huge; we run over 18,000 pods. At that scale, failure is inevitable, so you design your systems to be resilient.
When you operate at that scale, observability isn't a luxury. It is the eyes and ears of your infrastructure. But Digibee faced a common paradox in the tech world: the "Success Tax." As they grew, their tooling costs exploded. They were winning in the market, but their monitoring bill was punishing them for it.
The Wall of Cost vs. Visibility
Digibee previously relied on Datadog. While the platform functioned well, its pricing structure proved incompatible with Digibee's growth.
The cost became so prohibitive that the engineering team had to make impossible choices. To stay within budget, they could only deploy Application Performance Monitoring (APM) capabilities to a fraction of their fleet.
While metrics and logs were available across their infrastructure, full APM coverage was too costly to run at scale.
Talking about the APM at Datadog, because of the cost, we couldn't deploy it to the whole fleet... At the time it was like only 10% [visibility].
Imagine driving a sports car at 200 mph but only seeing through a small crack in the windshield. That was Digibee’s reality. Full APM coverage was limited to just 10% of their fleet. Developers were building and deploying code without understanding the full patterns of the applications they were creating.
They needed a change, but reliability could not be compromised.
The Dash0 Difference: Head-to-Head Performance
Tiago and his team evaluated alternatives carefully. As a high-performance platform, they could not afford to trade stability for savings.
Dash0 entered the picture as a robust, OpenTelemetry-native platform that matched their needs.
Datadog was super expensive and Dash0 was one third of the cost. It was an easy decision!
By leveraging Dash0’s efficient architecture and pricing model, Digibee unlocked the ability to monitor their entire fleet without breaking the bank.
Collaboration at the Speed of AI
Migrating observability tools can feel like replacing the engine of a plane mid-flight. It requires more than a vendor, it requires a partner.
Tiago highlights the responsiveness of the Dash0 team as a key factor. Support was so fast it became a running joke internally that responses felt automated.
This collaborative approach meant Digibee wasn’t just buying a tool; they were working alongside a team aligned with their OpenTelemetry-first mindset.
From 10% to 100%: A Revolution in Insight
The impact of switching to Dash0 was immediate and transformative. The previous limitations on APM coverage were removed.
The results spoke for themselves:
- Massive Coverage Increase: Digibee went from monitoring 10% of their applications to 100% coverage.
- Deep Context: Developers gained deep context on errors and performance bottlenecks, allowing them to spot patterns they had previously missed.
- Proactive Culture: The team established a weekly Wednesday meeting guided entirely by Dash0 alerts and signals.
Now developers can find the patterns and they're sent with a lot more deep context about what is happening.
With broader visibility across the engineering team, observability shifted from reactive firefighting to a more proactive approach to quality.
Conclusion: Observability Without Compromise
Digibee’s story proves that you don't have to choose between scale and cost, or between savings and performance. By embracing an OpenTelemetry-native approach with Dash0, they regained control of their infrastructure.
They expanded full APM visibility across their entire production environment, enabling their 15-person engineering team to operate with greater confidence.
