As companies build more software and adopt AI coding tools, the amount of data generated by their applications is growing rapidly. But more data does not necessarily mean better visibility. Excessive logs, incomplete instrumentation and unreliable telemetry can drive up infrastructure costs while making it harder for engineers to identify genuine problems. OllyGarden, a startup focused on improving telemetry quality, has raised $4 million to help businesses collect more useful data and make software monitoring more efficient.
The funding round includes Next Frontier Capital, Grand Ventures and ACTAI Ventures, alongside existing investor DIG Ventures. Observability companies Datadog, Grafana Labs and Dash0 also participated through strategic investments.
Launched in 2025, OllyGarden is developing technology that helps businesses improve the quality of telemetry data used to monitor software applications and infrastructure.
Tackling the Problem of Too Much Data
Observability tools help engineering teams understand how software behaves by collecting information such as logs, metrics and traces. This information can reveal errors, performance bottlenecks and system failures, allowing teams to investigate problems before they significantly affect users.
However, collecting and storing telemetry can become expensive as applications grow more complex. Poorly designed instrumentation can generate enormous volumes of repetitive or irrelevant information, while missing data can leave engineers without the context needed to diagnose an issue.
OllyGarden aims to address both problems by evaluating telemetry quality continuously. Its platform identifies unreliable or unnecessary data and helps teams correct instrumentation at its source, before the information reaches observability backends.
The approach is designed to reduce the amount of data companies need to process and store while improving the usefulness of the information that remains.
Turning Data Quality Into Action
Rather than simply identifying problems in dashboards, OllyGarden focuses on helping engineering teams implement fixes within their software development workflows.
Its technology is built around OpenTelemetry, an industry standard for collecting and managing telemetry across applications. By working with existing instrumentation practices, the platform aims to help organisations establish consistent standards across teams and code repositories.
According to OllyGarden, its work with customers has identified opportunities to reduce log volumes by as much as 85%. The company says these reductions can help lower observability costs while making important signals easier to identify.
The results are particularly relevant for organisations managing large software environments, where even small improvements in data collection can affect storage requirements, processing expenses and the time engineers spend investigating problems.
Introducing Minimum Viable Instrumentation
Alongside its funding announcement, OllyGarden introduced Minimum Viable Instrumentation, or MVI, a new capability for its AI OpenTelemetry Engineer, Rose.
Previously, Rose could examine existing instrumentation and recommend improvements. MVI extends those capabilities by identifying areas where instrumentation is missing and providing guidance on establishing a useful baseline.
The objective is to help companies collect enough information to understand their applications without instrumenting every possible event or generating unnecessary telemetry.
This capability also addresses a growing challenge associated with AI generated code. As developers use AI tools to produce software more quickly, instrumentation can become inconsistent, incomplete or excessively detailed.
MVI is designed to help teams apply common standards across codebases, pull requests and development workflows, making telemetry quality easier to maintain as software changes.
AI Agents and the Future of Observability
OllyGarden founder Juraci Paixão Kröhling argues that effective monitoring depends on the usefulness of collected information rather than its volume. The company wants to help engineers and AI agents work with reliable, actionable telemetry instead of overwhelming them with noise.
That distinction could become increasingly important as AI agents take on more software development and troubleshooting tasks. Automated systems need accurate context to identify failures and recommend appropriate fixes, making data quality a foundational requirement.
Funding the Next Stage of Growth
The $4 million investment will support OllyGarden’s technology development and its efforts to help more organisations improve telemetry quality.
With strategic backing from established observability companies, the startup is positioning itself within a market increasingly concerned with the cost and complexity of monitoring modern software.
Its central proposition is straightforward: companies should not have to choose between collecting enough information to understand their systems and controlling the cost of that information.
By combining continuous quality checks with AI assisted instrumentation, OllyGarden aims to make software monitoring more efficient, consistent and useful for both engineers and the AI systems increasingly working alongside them.
