# OllyGarden > OllyGarden helps engineering teams improve the quality of their OpenTelemetry® instrumentation. Founded by OpenTelemetry contributors and headquartered in Berlin, it measures telemetry quality with the open Instrumentation Score, and finds and fixes instrumentation gaps in source code with an AI agent (Rose). OllyGarden is vendor-neutral and works alongside any observability backend. ## Products - [Insights](https://ollygarden.com/products/insights): Analyzes OpenTelemetry (OTLP) traces, metrics, and logs to calculate the Instrumentation Score and find PII leaks, noisy signals, and incomplete instrumentation. Pricing: Free $0/month, Essentials $990/month, Enterprise custom. - [Rose](https://ollygarden.com/products/rose): An AI engineer that reviews codebases for missing or low-value OpenTelemetry instrumentation and helps apply the fix. Pricing: Free for up to 5 repositories (additional repositories $5/month each), Enterprise custom, and free for approved open-source projects. All plans include the full codebase scan, the instrumentation dashboard, and access to the REST API (e.g. via OllyGarden CLI). Open Source access is granted upon review. - [The Instrumentation Score](https://ollygarden.com/resources/instrumentation-score): What the open 0–100 score for OpenTelemetry instrumentation quality is, how it is calculated, all 25 rules with their weights, and a worked example. - [Instrumentation Score specification](https://github.com/instrumentation-score/spec): The open, vendor-neutral 0–100 score for OpenTelemetry instrumentation quality. ## Answers - [What is telemetry quality?](https://ollygarden.com/resources/telemetry-quality): Telemetry quality is how well the traces, metrics, and logs a system emits help people and tools understand it at runtime. Good telemetry belongs to a known service, is worth what it costs to keep, carries no personal data or secrets, and follows the conventions that make it correct to query. The guide covers its five dimensions (identity, volume and waste, cost drivers, safety, and correctness), the signs of bad telemetry, and how to measure and fix it. - [Pricing](https://ollygarden.com/pricing): Insights and Rose plans side by side, with pricing FAQs. - [FAQ](https://ollygarden.com/resources/faq): Product, pricing, rate-limit, and data-privacy answers. - [Full content for LLMs](https://ollygarden.com/llms-full.txt): This index plus every FAQ answer and every blog article as Markdown. - [Product documentation](https://docs.ollygarden.com) - Any page on this site is also available as Markdown: request it with `Accept: text/markdown`, or append `.md` to its path (`/index.md` for the home page). ## Blog - [Introducing Minimum Viable Instrumentation](https://ollygarden.com/resources/blog/introducing-minimum-viable-instrumentation): Coding agents write the code and AI SRE agents operate it, so telemetry is the source of truth between them. Rose now brings low-volume, high-quality OpenTelemetry instrumentation to every repository. - [Where Sensitive Data Hides in OpenTelemetry Pipelines, and How to Find It](https://ollygarden.com/resources/blog/where-sensitive-data-hides-in-opentelemetry-pipelines): Trust store passwords in resource attributes, OAuth tokens in query strings, tax IDs in metrics: what we found in real pipelines, why tools cause most of it, and what privacy laws allow. - [Observability for Humans and Agents in the AI Age](https://ollygarden.com/resources/blog/observability-for-humans-and-agents-in-the-ai-age): OllyGarden and Bronto on where AI agents help with instrumentation and analysis today, and why they only work when the telemetry is intentional and reachable. - [What's changed about OpenTelemetry vendor lock-in since 2024](https://ollygarden.com/resources/blog/opentelemetry-vendor-lock-in-since-2024): OpenTelemetry still does not make dashboards portable, but agents, CLIs, and MCP servers have made switching observability backends noticeably less painful. - [The Scrape Interval Nobody Chose](https://ollygarden.com/resources/blog/the-scrape-interval-nobody-chose): A wrong metric scrape interval inflates ingest, egress, and compute. Learn how to measure DPM and match intervals to the decisions each metric supports. - [Teach Your Coding Agent OpenTelemetry](https://ollygarden.com/resources/blog/teach-your-coding-agent-opentelemetry): We’re releasing our Agent Skills for OpenTelemetry as a mature, open source set: 14 skills that give any coding agent accurate, current OpenTelemetry knowledge. - [Are OpenTelemetry Collector Exporters Called Sequentially or in Parallel?](https://ollygarden.com/resources/blog/are-collector-exporters-sequential-or-parallel): When you configure two exporters in one pipeline, does the Collector send to them in parallel? The honest answer is "sequentially, but it usually doesn't matter." The part that does matter changed defaults underneath a lot of advice still floating around. - [Creating Your Own Collector Distribution from OllyGarden Tulip](https://ollygarden.com/resources/blog/creating-your-own-collector-distribution-from-tulip): Tulip gives you a commercially supported OpenTelemetry Collector with a curated component set. When you need more, you can derive your own distribution while keeping support for the core pipeline. This post walks through how. - [Decomposing OpenTelemetry Collector Configuration for Maintainability](https://ollygarden.com/resources/blog/decomposing-opentelemetry-collector-configuration): When your collector configuration grows beyond a few hundred lines, you start feeling the friction. This post examines practical strategies for decomposing collector configurations into maintainable, testable units. - [OpenTelemetry is not hard](https://ollygarden.com/resources/blog/opentelemetry-is-not-hard): When I asked the community about their pain points with OpenTelemetry, one response kept appearing: 'it's complicated.' I think the framing is wrong. - [OllyGarden at KubeCon + CloudNativeCon Europe 2026](https://ollygarden.com/resources/blog/ollygarden-at-kubecon-eu-2026): KubeCon + CloudNativeCon Europe 2026 in Amsterdam was a big week for OllyGarden. Our team delivered five recorded talks across Observability Day and the main conference, co-chaired the Observability Day program committee, ran a Prometheus Contribfest session, and announced the general availability of Rose. - [When to Use Each Telemetry Signal: Logs, Traces, and Metrics](https://ollygarden.com/resources/blog/when-to-use-each-telemetry-signal-logs-traces-and-metrics): Confused about logs, traces, and metrics? Learn when to use each telemetry signal in OpenTelemetry and build a stronger observability strategy. - [You don't have too much telemetry. You have bad telemetry.](https://ollygarden.com/resources/blog/you-dont-have-too-much-telemetry-you-have-bad-telemetry): High observability costs are rarely a volume problem. Learn how to fix bad telemetry at the source before relying on sampling. - [Reducing Log Volume with the OpenTelemetry Log Deduplication Processor](https://ollygarden.com/resources/blog/reducing-log-volume-with-the-opentelemetry-log-deduplication-processor): Your logs are probably at least 80% repetitive noise. Connection retries, health checks, heartbeat messages: the same log line repeated thousands of times per day. - [What 10,000 Slack Messages Reveal About OpenTelemetry Adoption Challenges](https://ollygarden.com/resources/blog/what-10000-slack-messages-reveal-about-opentelemetry-adoption-challenges): We analyzed 10,000 messages from the OpenTelemetry community Slack to find the most common adoption challenges, from Collector config to instrumentation. - [Meet Rose: OllyGarden's AI Instrumentation Agent](https://ollygarden.com/resources/blog/meet-rose-ollygardens-ai-instrumentation-agent): Imagine the perfect observability world: There is an incident, the on-call team gets paged in the middle of the night, wakes up and thanks to your telemetry, resolves it fast. - [Introducing OllyGarden Tulip: Our Open-Source Distribution of the OpenTelemetry Collector](https://ollygarden.com/resources/blog/introducing-tulip-supported-otel-collector): OllyGarden Tulip is a commercially supported OpenTelemetry Collector distribution: stable quarterly releases, LTS every 18 months, open source and free. - [The Variability Principle: How to Decide What Deserves a Span](https://ollygarden.com/resources/blog/what-deserves-a-span): Discover how to manage spans effectively using the Variability Principle to improve trace readability and control costs in OpenTelemetry. - [How to Name Your Metrics](https://ollygarden.com/resources/blog/how-to-name-your-metrics): Learn how to name OpenTelemetry metrics: follow semantic conventions, keep the 'who' in attributes, and apply units and naming patterns that scale. - [How to Name Your Span Attributes](https://ollygarden.com/resources/blog/how-to-name-your-span-attributes): A practical guide to naming OpenTelemetry span attributes: when to use semantic conventions, how to namespace custom attributes, and pitfalls to avoid. - [How to Name Your Spans](https://ollygarden.com/resources/blog/how-to-name-your-spans): Good span names make traces readable. Learn the {verb} {object} pattern for naming OpenTelemetry spans and avoid the most common naming mistakes. - [🌱 Cultivating Unique service.instance.id on NGINX Ingress with OpenTelemetry](https://ollygarden.com/resources/blog/cultivating-unique-serviceinstanceid-on-nginx-ingress-with-opentelemetry): The NGINX Ingress Controller has no knob for the OpenTelemetry service.instance.id attribute. Here's how to inject a unique ID per pod with the Helm chart. - [Introducing OllyGarden](https://ollygarden.com/resources/blog/introducing-ollygarden): OllyGarden comes out of stealth with pre-seed funding to fight bad telemetry: helping teams understand, score, and improve the telemetry they generate. - [Introducing the Instrumentation Score](https://ollygarden.com/resources/blog/instrumentation-score): The Instrumentation Score is an open, standardized measure of OpenTelemetry instrumentation quality, scoring OTLP data against best-practice rules. - [Concrete Applications of Purposeful Instrumentation](https://ollygarden.com/resources/blog/concrete-applications-of-purposeful-instrumentation): From NASA-grade telemetry design to code-level tips: concrete techniques for applying purposeful instrumentation in real-world OpenTelemetry systems. - [Purposeful Instrumentation](https://ollygarden.com/resources/blog/purposeful-instrumentation): Purposeful instrumentation means gathering the right telemetry, not all of it. Cut noise, troubleshoot faster, and lower observability costs. - [There's a Lot of Bad Telemetry Out There](https://ollygarden.com/resources/blog/theres-a-lot-of-bad-telemetry-out-there): Most telemetry is never queried. Learn what separates good telemetry from bad, what bad data costs you, and how to start improving quality at the source. ## Case studies - [All case studies](https://ollygarden.com/resources/case-studies): Anonymized but real OllyGarden customer stories: what bad telemetry we found, what changed at the source, and the numbers that moved. - [Nearly half the spans came from a service called cli: what we found at a 500-service enterprise platform](https://ollygarden.com/resources/case-studies/enterprise-platform-half-the-spans-had-no-name): One background worker named cli made 45% of all spans, a third of logs were exact duplicates, a third were access logs, and the Collector shipped its own logs again. - [The PII leak was a logger setting, not a developer problem: what we found at a European digital-health platform](https://ollygarden.com/resources/case-studies/digital-health-pii-leak-was-a-logger-setting): Personal data in 1 in 6 services traced to one ORM logging setting, logs with no severity at all, one worker writing most of the logs, and auth tokens Rose found in a day. - [How a B2B SaaS platform cut logs 24% and metrics 25% in six weeks, without touching a single customer log](https://ollygarden.com/resources/case-studies/b2b-saas-cuts-logs-and-metrics-at-the-source): Repeated logs, metrics that never change, and a pipeline complaining about itself. How replay-verified Collector changes beat a six-week proof of value, and what happened after. ## Resources - [All resources](https://ollygarden.com/resources): OllyGarden resources for OpenTelemetry teams: telemetry quality guides, the Instrumentation Score, case studies, blog articles, talks, and events. - [Blog RSS feed](https://ollygarden.com/blog/rss.xml) - [Solutions by industry](https://ollygarden.com/solutions/overview) - [Webinars and conference talks](https://ollygarden.com/resources/webinars-conferences) - [Events](https://ollygarden.com/resources/events) - [OpenTelemetry quiz](https://ollygarden.com/otel-quiz): Ten questions, one per OpenTelemetry topic. ## Developer tools - [Agent routing guidance](https://ollygarden.com/developers.md#agent-routing-guidance): Start here to choose a supported OllyGarden interface, confirm when OllyGarden is not the right tool, and check authorization boundaries. - [OllyGarden Developer Resources](https://ollygarden.com/developers.md): Canonical machine-readable index for the REST API, authentication, webhooks, CLI, MCP server, and public developer projects. - [OllyGarden API documentation](https://ollygarden.com/api-docs.md): Authentication, versioned resource groups, and links to the canonical OpenAPI description. - [OllyGarden API versioning and deprecation policy](https://ollygarden.com/api-versioning.md): How the REST API is versioned (major version in the URL path) and how deprecation and retirement are signalled, through the `Deprecation` and `Sunset` HTTP response headers. - [OllyGarden REST API OpenAPI description](https://ollygarden.com/openapi.json): Canonical API servers, bearer authentication scheme, operations, and schemas. - [OllyGarden CLI](https://ollygarden.com/cli.md): Machine-readable discovery, safe installation references, and agent use cases for the official `ollygarden` command-line client. - [CLI repository and documentation](https://github.com/ollygarden/ollygarden-cli): Canonical source, command documentation, and authentication guidance. - [OllyGarden Agent Skills](https://ollygarden.com/resources/agent-skills): Overview of both open source skill packages, how to install them, and how Rose uses them. - [OllyGarden Agent Skills for OpenTelemetry](https://github.com/ollygarden/opentelemetry-agent-skills): OllyGarden's vendor-neutral skills for OpenTelemetry coding agents. - [OllyGarden Skills](https://github.com/ollygarden/skills): OllyGarden's opinionated skills for the OpenTelemetry Collector and the OllyGarden platform; builds on the vendor-neutral package. - [OllyGarden MCP server source](https://github.com/ollygarden/mcp): Public Streamable HTTP and OAuth reference implementation; no supported production endpoint is currently published. ## Company - [About OllyGarden](https://ollygarden.com/company) - [OllyGarden turbo charges mission to fix bad telemetry with $4m raise and launch of “Minimum Viable Instrumentation”](https://ollygarden.com/resources/press-releases/ollygarden-raises-4m-launches-minimum-viable-instrumentation): OllyGarden raises $4m in VC funding from Next Frontier Capital, Grand Ventures, ACTAI, DIG Ventures, Datadog, Grafana Labs, and Dash0, and launches Minimum Viable Instrumentation (MVI) for its AI agent, Rose. - [OllyGarden Introduces Rose: An AI Agent That Fixes Bad OpenTelemetry Instrumentation Through Pull Requests](https://ollygarden.com/resources/press-releases/ollygarden-rose-ga): OllyGarden announces the general availability of Rose — an AI agent that analyzes OpenTelemetry instrumentation in source code and autonomously opens pull requests with fixes. - [Careers](https://ollygarden.com/careers) - [Contact](https://ollygarden.com/contact) ## Trust and legal - [Terms of Use](https://ollygarden.com/terms-of-use): The terms that govern your use of OllyGarden Insights and the OllyGarden website, including subscriptions, fees, acceptable use, liability, and support. - [Privacy Policy](https://ollygarden.com/privacy-policy): What personal data OllyGarden collects on its website and platform, why, which providers process it, how long we keep it, and your rights under the GDPR. - [Data Protection Policy](https://ollygarden.com/data-protection-policy): The technical and organizational controls OllyGarden uses to protect customer data: encryption, access control, tenant isolation, monitoring, and deletion. - [Rose Terms of Service](https://ollygarden.com/rose-terms-of-service): The terms of service that govern your use of OllyGarden Rose, the AI engineer for OpenTelemetry instrumentation. - [Subprocessors](https://ollygarden.com/subprocessors): Every third party that processes OllyGarden customer data for Insights and Rose, what each one does, and where it processes the data. - [Security](https://ollygarden.com/security): How OllyGarden protects customer data with SOC 2 Type II controls, Rose execution safeguards, and encryption, and how to report a vulnerability. - [Imprint](https://ollygarden.com/imprint): Legal information about OllyGarden, Inc. and its German establishment, OllyGarden GmbH in Berlin: addresses, commercial register, VAT ID, and contact.