Empirik is entering the market with a focused pitch: infrastructure teams should not have to wait for systems to fail before they understand what went wrong. The startup, incubated by Sequoia Capital, announced Tuesday that it is spinning out as an independent company after raising $21 million in seed funding from Sequoia, Canapi, and Alumni Ventures.
The company is building an AI observability tool designed to track system changes and infer how those changes could affect infrastructure. Its broader goal is to help teams prevent and resolve incidents by understanding risk before an outage occurs.
How Empirik Started
Empirik grew out of work by Avon Puri and Sudheer Dhurjati, two technology leaders with deep infrastructure experience. Before joining Sequoia Capital in 2020 as chief digital and information officer, Puri spent over a decade running infrastructure at Rubrik and VMware.
Three years ago, as large language models began showing their true potential, Puri and Dhurjati saw a practical opportunity for AI inside infrastructure operations. Their idea was not simply to make alerting faster. It was to shift the work from reacting after failures to predicting where failures might happen.
That thinking became Empirik. The product is built to watch changes across systems and reason about their possible ripple effects. In large environments, where many services, workflows, and dependencies interact, a single change can matter far beyond the place where it was made.
Sequoia incubated the startup in 2023. Earlier this year, the firm recruited Kartik Chandrayana, formerly Quantum Metric CPO and Salesforce observability VP, to serve as CEO.
What The Product Is Trying To Solve
Modern infrastructure teams already spend significant time trying to keep systems available. Sequoia partner Bogomil Balkansky told TechCrunch, “There has always been a lot of money spent in keeping systems up and running,” while arguing that most existing observability tools do not fully understand complex system dependencies.
Empirik is positioned around that gap. Rather than looking only at symptoms after a problem appears, the software is meant to evaluate changes before they become incidents. The company’s central premise is that infrastructure risk often comes from the relationship between systems, not from a change in isolation.
In practical terms, the product acts as a kind of autonomous infrastructure engineer. It can support routine troubleshooting, assess system updates, and help decide when a change needs closer review.
According to Balkansky, Empirik functions like an autonomous “traffic cop” for infrastructure work. That means it can permit low-risk changes, set guardrails around larger changes, and flag the most dangerous updates for human review.
Why AI Matters Here
Empirik is arriving at a moment when AI is changing how quickly software gets written. Chandrayana told TechCrunch that as AI accelerates the pace of software development, tools that help infrastructure engineers keep up with constant system changes are becoming more important.
That creates a simple operational challenge. If software teams can ship faster, infrastructure teams must understand changes faster as well. Otherwise, the gap between development speed and operational confidence can widen.
Empirik’s answer is to apply agentic AI to infrastructure engineering. Chandrayana summarized the ambition this way: “What agentic AI did for software, Empirik wants to do for infrastructure engineering.”
The company also compares its goal to the effect that Cursor and Claude Code have had for software developers. The point is not to replace engineers, but to automate certain tasks so technical teams can move significantly faster.
For DevOps and site reliability engineering teams, that could mean less time spent on routine investigation and more time focused on higher-value priorities. The value proposition depends on whether the product can reliably understand infrastructure relationships and make useful calls about risk.
Early Customers And Market Position
Empirik says it has already brought on customers since launching earlier this year. Those customers range from startups to several Fortune 500 players, including S&P Global, Guardant Health, and a major consumer packaged goods (CPG) company.
That early customer mix matters because the product is aimed at environments where change is frequent and dependencies can be difficult to follow manually. Startups may move quickly, while large companies often have more complex systems and stricter expectations around availability.
On competition, Balkansky claims Empirik is in a category of its own for now. He described it as a complementary layer to AI SRE platforms such as Resolve and Sequoia-backed Traversal.
That framing is important. Empirik is not being presented simply as another dashboard or monitoring tool. It is being positioned as a layer that understands changes, dependencies, and risk across massive environments.
The Bigger Implication
The core bet behind Empirik is that infrastructure work can become more predictive. Traditional observability helps teams see what is happening. Empirik wants to help teams understand what may happen next when systems change.
If that approach works, it could alter how engineering teams think about incident prevention. Instead of treating outages as events that trigger investigation, teams could treat risky changes as events that trigger prevention.
The company still has to prove that its AI can handle the complexity of real production environments. But its launch shows where infrastructure tooling is heading: toward systems that do more than observe, and toward AI that helps engineers act before failure becomes visible.