Tsuga raises $35M for AI agent observability without per-byte pricing

Paris-based startup Tsuga has raised $35M in a Series A round, just six months after emerging from stealth. The company is building an observability platform specifically for the era of AI agents, where telemetry volumes are growing faster than existing monitoring tools can handle.

Who is behind Tsuga

Two former Datadog employees founded the startup. Datadog effectively created the modern observability market and is currently valued at roughly $40B. Datadog alumni send a signal in the industry: they know the market from the inside and see an unfilled niche in it.

Tsuga came out of stealth less than a year ago and has already closed a sizable round. That suggests investors believe the niche is real, the timing is right, and the company needs money right now so it does not miss the window.

Why the old model breaks

Standard billing in observability is payment for every byte of telemetry. For classic web applications this worked: a service generates predictable volumes of logs and metrics, and cost scales linearly.

AI agents wreck that math. In a single session, an agent orchestrator creates hundreds of nested tool calls, prompts, model responses and traces. Each one is needed for debugging, but together they produce volumes at which per-byte billing becomes financially untenable.

• AI agents produce orders of magnitude more telemetry than ordinary services
• Per-byte billing makes full monitoring impractical
• Data sampling hides critical errors in call chains
• Existing tools are not optimized for AI tracing patterns
• Companies have to choose between data completeness and budget

Tsuga wants to offer a different economic model: not per byte, but based on another metric that keeps full monitoring affordable even as workloads grow explosively.

Inside the customer's cloud

The startup's second key choice is its deployment model. Tsuga installs inside the customer's own infrastructure rather than pulling telemetry data onto external servers.

For the enterprise market, this is critical. AI agent traces can contain prompt contents, fragments of processed documents and user data. Handing all of that to a third-party vendor creates legal and regulatory risk, especially in finance, healthcare and the public sector. The "in your cloud" model is not just an architectural choice; it is an argument in deals with large businesses.

What this means

The observability market for AI agents is only taking shape, and the first serious money has already arrived. With founders who have Datadog experience, a rejection of the per-byte trap and deployment inside the customer's infrastructure, Tsuga enters with a well-considered set of arguments.

If AI agents become mass-market infrastructure in the next two to three years, the tools to monitor them will become a mandatory layer of the stack, and a race for this market is already underway.