Origins of the Predictive Infrastructure Model
Avon Puri spent over a decade managing infrastructure at major firms like VMware and Rubrik. During his tenure at Sequoia Capital, which began in 2020 as chief digital and information officer, he observed how large language models began to change the way software gets written. He teamed up with fellow Sequoia IT lead Sudheer Dhurjati to address a chronic problem in the industry. They wanted to shift the focus from reactive incident response to proactive outage prevention.
Most current systems rely on alerts triggered after a failure has already disrupted service. The pair realized that tracking system changes in real time could allow for the prediction of failures before they occur. This vision became the foundation for Empirik, a startup designed to map system dependencies and catch ripple effects across complex digital environments. The company acts as an autonomous layer that monitors the health of infrastructure as developers roll out updates.
Growth and Strategic Funding
Sequoia Capital incubated the startup in 2023. They identified the need for a dedicated observability tool that could handle the complexity of modern cloud environments. To lead the transition to an independent entity, the firm brought in Kartik Chandrayana, who previously served as Chief Product Officer at Quantum Metric and VP of observability at Salesforce. His appointment signaled the shift from an internal project to a market-ready business.
Empirik emerged from stealth on Tuesday with $21 million in seed funding. The round included participation from Sequoia, Canapi, and Alumni Ventures. This influx of capital supports the development of a platform that functions as an automated traffic controller for DevOps teams. It manages the risk associated with changes by allowing low-risk updates to pass while requiring human intervention for more sensitive deployments.
Industry Impact and Future Direction
Bogomil Balkansky, a partner at Sequoia, notes that a massive portion of IT budgets goes toward basic system maintenance and uptime. Existing tools often lack the capacity to track the connections between various system components, leading to blind spots. Empirik differentiates itself by acting as an autonomous infrastructure engineer, freeing human teams to focus on core product innovation rather than routine troubleshooting.
Early adoption confirms the demand for this approach. The company already counts several Fortune 500 businesses among its clients, including Guardant Health. These firms use the platform to maintain system stability even as the speed of software deployment increases. The goal is to match the efficiency gains that coding assistants like Cursor and Claude Code brought to the software development process.
The Changing Landscape of Site Reliability
As AI tools accelerate the development cycle, the pressure on infrastructure teams intensifies. Changes happen at a pace that manual review processes cannot easily support. Empirik positions itself as a critical layer that keeps pace with these rapid deployments. Chandrayana views the product as the infrastructure equivalent of agentic AI for software development, providing the guardrails necessary for modern digital architecture.
While other platforms exist in the site reliability space, such as Resolve and the Sequoia-backed Traversal, Empirik aims to carve out a specific niche. It functions as a complementary layer rather than a total replacement for existing observability suites. The company enters a competitive field with significant backing and a clear focus on automating the most dangerous part of infrastructure maintenance: the deployment of unknown system changes. Monitoring whether this model successfully scales across diverse, legacy enterprise environments will be the next major benchmark for the firm.