A Sequoia Capital-backed startup wants to do for IT teams what tools like Cursor already did for coders — let them work faster by handling the boring, risky parts of the job automatically.
Tech outages are expensive. They cost companies money, trust, and hours of stressed-out engineering time. A new startup called Empirik says it has built an AI tool that can catch these problems before they turn into a full-blown outage — not after.
Empirik just announced that it raised $21 million in seed funding, and it’s officially spinning out as an independent company after being built inside Sequoia Capital.
What Does Empirik Actually Do?
In simple terms, Empirik watches everything that changes inside a company’s tech systems — a new code push, a config update, a server change — and predicts what might break because of it.
Instead of engineers finding out something is broken after customers start complaining, Empirik tries to flag the risky change beforehand.
The founders describe it almost like a traffic controller for IT systems:
- Low-risk changes get a green light and go through automatically.
- Medium-risk changes get extra guardrails and monitoring.
- High-risk changes get flagged for a human engineer to review before anything breaks.
This matters because most IT teams today are stuck being reactive — they fix things after they fail. Empirik is trying to flip that model so teams catch problems earlier, similar to how modern AI observability tools are reshaping DevOps and site reliability engineering (SRE) work.
Who’s Behind It?
Empirik was built inside Sequoia Capital by two people who know infrastructure well:
- Avon Puri — Sequoia’s Chief Digital and Information Officer, who previously ran infrastructure at Rubrik and VMware.
- Sudheer Dhurjati — another Sequoia IT leader who worked alongside Puri.
The two noticed something simple but important: as AI made software development faster, IT infrastructure teams were struggling to keep up with the pace of change. So they built a tool to help.
Sequoia later brought in Kartik Chandrayana, a former CPO at Quantum Metric and observability VP at Salesforce, to run the company as CEO.
The $21 million seed round came from Sequoia Capital, Canapi, and Alumni Ventures.
Early Customers Already Onboard
Even though Empirik only launched publicly this year, it says it has already signed up customers ranging from early-stage startups to large enterprises — including names like S&P Global, Guardant Health, and a major consumer packaged goods company.
That’s a strong early signal. Big companies don’t usually adopt brand-new infrastructure tools unless they’re solving a real, expensive problem.
Why This Matters for the Bigger AI Startup Trend
Empirik is part of a growing wave of AI infrastructure startups trying to automate work that used to require a human expert watching dashboards all day. Just like AI coding tools changed how software gets written, tools like Empirik are trying to change how IT systems get maintained.
The company says it isn’t really competing head-on with existing AI SRE (site reliability engineering) platforms like Resolve or Sequoia-backed Traversal. Instead, it sees itself as working alongside those tools — focusing specifically on predicting and preventing risky changes before they cause damage.
The Bottom Line
Empirik’s bet is simple: most companies spend huge amounts of money keeping their systems running, and most existing tools only help after something has already gone wrong. If Empirik’s AI can actually predict outages before they happen, it could save engineering teams a lot of time, stress, and money.
With $21 million in fresh funding and real enterprise customers already testing the product, Empirik is one of the more interesting AI startup funding stories to watch this year.
Learn more:
