A wave of startups challenging database heavyweights like Snowflake, Databricks, and Datadog is experiencing explosive revenue growth, fueled by the widespread adoption of AI agents. This momentum could trigger a flurry of startup transactions, spanning new funding rounds and mergers and acquisitions.
Earlier this Tuesday, I reported that ClickHouse's annual recurring revenue hit a new high. This company, which offers free open-source software and paid cloud services to help businesses analyze massive datasets, saw its product usage by OpenAI climb significantly this year. OpenAI's move is quite telling: the ChatGPT developer is also a customer of the publicly listed Datadog, a leading observability firm that monitors various activities of both humans and AI agents on computers.
According to sources, ClickHouse's last funding round took place in January at a valuation of $15 billion, and it has already received acquisition inquiries from multiple strategic buyers this year. CEO Aaron Katz told me, "We won't ignore external acquisition interest since we have a responsibility to shareholders, but we have no plans to sell the company. On the contrary, we intend to continue developing independently."
ClickHouse is not the only beneficiary of the massive data demands brought by AI agents. Cribl, which specializes in telemetry technology to help enterprises migrate and store data for monitoring purposes, says its annual recurring revenue is approaching $400 million, a 33% increase since February. "Our enterprise clients are facing explosive data growth, partly driven by the rise of AI and agents, and we're helping them navigate this situation," said Clint Sharp, co-founder and CEO of Cribl.
The eight-year-old company serves clients like Zoom, ServiceNow, and Hilton, helping them route data to lower-cost storage while matching the most efficient monitoring tools. Sharp noted, "The real pain point we solve is that enterprise data continues to grow at a compound rate of 30%, yet IT budgets aren't growing in tandem." Cribl's last venture round was in August 2024 at a valuation of $3.5 billion. Sharp revealed that the company plans to launch an IPO within the next two to three years.
Grafana, a 12-year-old company whose tools let developers visualize the operational status of applications, cloud servers, and AI agents in real time, has also seen accelerating revenue recently. Chief Marketing Officer Scott Fingerhut attributes some of this growth to its own AI assistant, which helps enterprises monitor the operational activities of their applications. Fingerhut noted that half of Grafana's customers used this AI assistant this year to configure system monitoring and debug tools. The company surpassed $400 million in annual recurring revenue last September.
Banking software industry insiders say Grafana is often viewed as a potential acquisition target for large enterprises, with major players eyeing the data monitoring sector likely to make a move. Large corporations are already snapping up startups focused on data monitoring. Publicly listed Elastic, which provides an open-source big data search engine, acquired Deductive AI on Monday for $85 million, a product that helps engineers monitor data and troubleshoot issues. Two weeks ago, Dynatrace spent $915 million to acquire Arize AI, which specializes in tracking the performance of AI models and agents.
Snowflake completed its $1 billion acquisition of Observe in June, betting that customers who have stored data on its platform will need tools to view application performance. Prior to this deal, cybersecurity firm Palo Alto Networks acquired observability startup Chronosphere for $3.35 billion. Investors and bankers predict more acquisitions will follow.
Cybersecurity vendors, tech giants like Google, Amazon, and Microsoft, as well as network equipment makers like Cisco, are all potential buyers of data monitoring startups. Mike Jung, a partner at FoundersCircleCapital, which has invested in Databricks, Vercel, ClickHouse, and Cribl, said, "As AI agents take on more autonomous work, enterprises need visibility into what these agents are actually doing, along with guardrails to intercept failures and security risks before they cause real damage."