We’re excited to announce our partnership with ClickHouse. Herald is an AI DevOps agent in House Mates, ClickHouse's partner program, and our ClickStack and ClickHouse Cloud integration is live. ClickHouse customers get up to $25,000 in free Herald usage.
ClickHouse is the telemetry engine. Herald is the intelligence layer. Together, customers can keep more, see more, and fix issues faster. You can learn more and get started here.
Herald’s Production Intelligence Natively Complements ClickHouse’s Speed and Efficiency
ClickHouse is becoming the storage engine of choice for modern observability. The combination of columnar storage, efficient compression, and real-time ingestion makes it a faster, more scalable, and (most importantly) a far more cost-effective option than traditional observability systems.
As software systems scale, cost-efficient and scalable storage is critical. Otherwise, observability costs climb as fast as your data volume. ClickStack's columnar storage and compression cut that, reducing storage for OpenTelemetry data by up to 90 percent. Just as important is the ability to interpret that data quickly. When production breaks, finding the right data at the right time determines how fast a team resolves the outage. And the more data your systems generate, the harder it becomes for engineers to quickly find what they need.
Herald brings that interpretation layer. It joined House Mates with a production-ready ClickStack and ClickHouse Cloud integration. With Herald, engineering teams who are already seeing performance and cost savings from ClickHouse will be able to understand their infrastructure better, efficiently debug production outages, and even get predictive warnings about potentially brewing incidents.
Herald on ClickHouse gives your team three things:
One store, one map.
ClickStack unifies your logs, metrics, traces, and session replays. Herald builds a context graph on top: your schemas, your service dependencies, and an understanding of what normal looks like linked directly to your code and infra. When API latency spikes after a code deployment, Herald knows which service is tied to the spike and quickly pinpoints the deployment that led to the regression. No more scrolling through countless dashboards, logs, and git commits to figure out what went wrong. Herald does it for you, fast.
Keep more, catch more.
ClickHouse's efficient storage enables you to gather large volumes of telemetry. Without the right agent in place, that data can increase noise and further bury signal. Herald, on the other hand, monitors the health of your production system, and tells you as soon as things start going wrong: The agent detects anomalies, investigates multiple possible root causes in parallel, filters out false positives, and accurately surfaces root cause. What reaches you is a validated incident, well before your incident management platform alerts on it.
Ask anything, anytime.
ClickHouse’s separation of storage and compute means an investigation can be exhaustive and an engineer can be curious. Your team simply asks Herald a question, just as they would ask a teammate, and they get an immediate answer drawn from your telemetry, code, developer tools, and team docs. How is this endpoint performing? What are the upstream and downstream dependencies of this service? What’s the standard Terraform config for new infra? Answers come back grounded in your data, with Herald’s understanding of your product, code, infrastructure, and business as the basis.
"Customers consolidate on ClickHouse because it’s the most performant, scalable, and cost-effective platform to store and access their data. Now Herald is putting that speed to work by bringing AI DevOps intelligence and action to ClickHouse customers. Herald helps catch issues earlier, find root cause in minutes, and get more value out of the telemetry they already trust."
— Abhinav Mehla, VP, Global Partners and Business Development, ClickHouse
Access is easy. Understanding is the hard part.
Any tool can query ClickHouse. Herald's difference is the context graph it builds at onboarding, which is why it goes straight to the right data instead of rediscovering your environment on every query. This makes each query more specific and relevant, which enables fast, accurate responses. Herald is also built for secure production environments: The agent is read-only, writes fine-grained queries to avoid egressing large volumes of data, and analyzes telemetry query results without storing them. This precision is only possible because Herald knows what it needs to do up front, rather than exploring broadly on every query. You can learn more at herald.dev/clickhouse.
“ClickHouse changed the economics of observability, and that makes this next step possible. Herald is the intelligence layer that further harnesses the power of ClickHouse by understanding your production environment and driving reasoning, analysis, and answers for you. With ClickHouse and Herald, engineering teams can get more value and insights out of their data, faster, and at scale.”
— Vikram Sreekanti, Co-founder and CEO, Herald
Up to $25,000 in free Herald usage for each ClickHouse customer
Every ClickHouse customer can receive up to $25,000 in free Herald usage for predictive issue detection, on-demand root cause analysis, and technical answers across their telemetry, code, and infrastructure.
Claim your $25,000 in free Herald usage for ClickHouse here. Or simply try Herald in your terminal now for free to see how it works: npm install -g @herald-ai/herald