Grafana's growth marks AI transition from experimentation to real-world deployment

Grafana Labs, a 12-year-old observability company, has surpassed $600 million in annual recurring revenue and 10,000 global customers, marking a significant milestone in the transition of AI from experimentation to real-world deployment. The company's CEO, Raj Dutt, stated that businesses are moving beyond proving AI's value to addressing practical concerns such as cost, governance, performance, and reliability. Grafana's growth is attributed to AI, which has driven increased business from AI companies and higher usage of its AI products by existing customers. AI has also led to more complex observability challenges, generating larger data volumes and new issues to monitor. Traditional monitoring focused on outages and slowdowns, but AI agents can now behave unpredictably, such as recommending competitors or creating security risks. Grafana's Adaptive Telemetry tools aim to reduce the amount of data customers pay to monitor, despite costing the company nearly $100 million in potential revenue. Dutt noted that the revenue quality is better because customers pay more closely aligned with their actual usage. In July, Grafana launched six new AI features, including its Agent Observability product, which is used by hundreds of customers to monitor AI agents, tracking speed and reliability. Over 18,000 organizations, including Alter Domus and Deutsche Telekom, use Graf, which is described as the company's fastest-growing product. Dutt emphasized that the majority of Grafana Assistant's users are paying customers. Grafana competes with companies like Datadog, which reported that its customers are increasingly deploying AI and using its product to monitor and secure it, according to cofounder and CEO Olivier Pomel.