Ask a CTO to name what keeps them up at night and you’ll hear a familiar story: 3 AM system-down alerts, multi-million dollar invoices from the very vendors that were supposed to prevent them, and nightmares about the next surprise their infrastructure is quietly plotting. This is the paradox of observability: the tools meant to bring clarity to mission-critical software environments by tracking system health and troubleshooting failures have become a source of immense technical, operational and financial risk. 

Gabriel-James Safar and Sébastien Deprez, co-founders of Tsuga, have lived the software observability crisis from the inside. After their first startup was acquired by Datadog, they spent over five years building the core monitoring products inside one of the industry’s most defining companies. They experienced observability at true hyperscale: its breakthroughs, its exponential growth, and its fundamental limits.

Together with a talented team, they’re bringing order to the chaos of modern software development, which is why we’re leading Tsuga’s seed round.

A New Foundation for Observability 

Even before the AI wave, the observability paradigm was already under strain. Engineering leaders faced an impossible choice between creating massive operational overhead to run in-house open-source stacks or absorbing the cost of third-party infrastructure that could reintroduce data blind spots.

Now, the shift towards AI-native software development is breaking it completely. Enterprises that are building for the AI era urgently require a new foundation for their observability infrastructure.   

With Tsuga, they’re reimagining observability, empowering teams to focus on their mission rather than their dashboards. The platform runs entirely within a customer’s own cloud environment, giving enterprises contextual understanding of their systems. This architecture forms the foundation for causal and agentic reasoning: observability that acts intelligently to prevent and resolve issues with speed.