ClickHouse and Hud Put Production Behavior at the Center of AI Software Development
For years, software teams have focused on improving how quickly they can write, test, and deploy code. The rise of coding agents is changing that equation. When AI can generate or assist with a growing share of software, the critical question becomes what engineering teams can learn about those changes once they meet real production conditions.
Hud says AI now generates or assists with 42% of the code developers ship, and that percentage is expected to reach 65% by 2027. The company argues that as development accelerates, production context becomes increasingly important for deciding which changes should move forward and which require closer examination.
That is the focus of a new integration between Hud and ClickHouse. By connecting Hud's Runtime Code Sensor with ClickStack, ClickHouse's open-source observability stack, the companies are bringing operational and code-level information into the same development workflow.
Observability Becomes an Engineering Input
The integration is built around two complementary views of an application.
ClickStack gives engineering teams broad visibility into applications and infrastructure, helping identify the service, deployment, or endpoint connected to an issue. Hud operates closer to the code, following function-level behavior and connecting production activity with the changes responsible for it.
Shared trace IDs allow the two platforms to be connected by a coding agent. An issue identified through ClickStack can therefore lead directly to the relevant code-level context in Hud.
"Our users already trust ClickHouse to store and query their Open Telemetry data at scale," said Mike Shi, Head of Observability of ClickHouse. "The shift now underway is from simply watching systems or investigating issues to using that data to make day-to-day engineering decisions. Hud connects observability data to the code and changes behind it, making the entire stack more useful for teams building and shipping software with AI."
The distinction matters because observability has traditionally been associated with understanding what happened in production. The integration is designed to connect that information with decisions about code changes themselves.
A Feedback Loop for AI-Generated Changes
Hud and ClickHouse are also positioning runtime information as something engineering teams can use before a change is deployed.
The Hud+ClickStack integration supports pre-deployment risk assessment for code changes, allowing teams to evaluate changes against real-time information about how affected code behaves in production. Higher-risk changes can be held for additional review and deeper context, while safer changes can move faster or be automatically merged.
That approach creates a feedback loop between production and development. Instead of AI-generated code moving through a process based primarily on what the change looks like, teams can incorporate information about how related code behaves under real production conditions.
"AI is accelerating how quickly teams can generate code, but safely shipping it with high confidence requires production context," said May Walter, CTO of Hud. "Hud and ClickHouse bring real production behavior at an unparalleled breadth and depth, so teams can build a production-aware AI SDLC: gating changes before they ship, proactively verifying them once deployed, and fixing issues as they arise - all using real runtime truth. Together with ClickHouse, we are bringing that intelligence across the entire AI SDLC."
When Deployment Reveals the Problem
The same connection can be used after code reaches production.
Many issues can begin with relatively small changes. A query can slow down, a code path can consume more resources, or a function can behave unexpectedly under a particular workload. Hud is designed to identify such issues at the function level and provide forensic context, while ClickStack contributes the wider operational picture.
The integration supports release verification, automatic reversion upon regression, automated detection and investigation, and agentic workflows that can create pull requests addressing underlying code issues. It also supports rollback and remediation workflows.
The objective is to reduce the distance between identifying a production problem and understanding the code responsible for it.
Why Runtime Context Matters
The companies' approach reflects a shift in the role of production data within the development lifecycle. Coding agents can already work with source code and other development inputs; the integration is designed to give those workflows access to runtime operational data, function-level context, and application behavior.
"Like every modern engineering organization, a growing share of our code is now written with AI," said Rom Kadria, Senior Software Engineer, monday.com. "We write code much faster, but the challenge has shifted to shipping just as quickly while maintaining confidence that new code won't cause harm. ClickHouse gives us the wide operational picture at scale, while Hud gives us the runtime intelligence and next-level introspection needed to evaluate and ship AI-generated code confidently. When issues do arise, combining ClickHouse and Hud allows us to triage and resolve them quickly. For a company building with AI, that combination is the obvious choice."
Bringing Production Into the AI SDLC
Engineering teams can begin using the integration by installing the Hud SDK and connecting it to their ClickStack service. Hud's runtime intelligence then flows alongside the OpenTelemetry data teams already collect.
The larger idea is to make production behavior part of the AI development loop rather than treating it solely as something engineers examine after deployment. With ClickHouse providing the broader operational picture and Hud connecting that picture to code-level behavior, the integration is designed to help teams assess changes, verify releases, and respond to problems using the same underlying source of truth: what the software is actually doing in production.
