Coralogix has raised $200 million in Series F funding at a $1.6 billion post-money valuation, just 11 months after its $115 million Series E, as reported by TechCrunch. The round was led by Advent and the Canada Pension Plan Investment Board, with Greenfield Partners and Brighton Park Capital participating. Total raised to date: $550 million.
The Boston-headquartered, Israel-founded observability company sells monitoring infrastructure: logs, metrics, and traces that tell engineering teams what their software is doing and why it broke. It competes with Datadog, New Relic, and Splunk. The thesis behind the raise is that autonomous AI agents in production create a new monitoring problem. The more software that runs without a human in the loop, the more you need to know when it goes wrong.
The number that matters most is buried in the CEO interview. More than half of Coralogix enterprise customers now investigate incidents through the company's AI agent, Olly, or through their own LLMs connected via CLI and MCP interfaces, rather than logging into a dashboard. CEO Ariel Assaraf put it directly: "The interface layer is slowly getting eroded." Customers want to ask an AI assistant what's wrong, not click through panels.
The financials underneath: 60%+ revenue growth over the past year, roughly 30 customers spending more than $1 million annually, $100M+ ARR crossed over a year ago, and 5,000+ customers including IBM, Tradeweb, and JFrog. The company says it raised to accelerate, not to extend runway, and is preparing to operate with public-company discipline.
Enterprise engineering teams are voting with their workflows, and the vote is against the dashboard. When more than half of a 5,000-customer base stops opening the UI and starts querying through an LLM or CLI, the buyer is telling the market that the interface is no longer where value is judged. Value is judged at the data layer and the agent layer. The thing buyers will pay $1M+ for is not a prettier panel. It is a system their agents can interrogate and a system that can watch their other agents.
There is a second signal: enterprises are deploying autonomous systems they do not fully trust. The entire pitch rests on the gap between "we shipped AI agents into production" and "we can prove what those agents did." That gap is now a procurement line item.
If you sell software with a UI, your UI is depreciating as a moat. The defensible surface is moving to two places at once: the structured data an agent can call, and the trust artifacts that prove your agentic features are safe. A vendor still investing primarily in dashboard polish in 2026 is investing in the layer that is eroding. The vendors winning this round are the ones who made their product invokable by an agent and made their monitoring the trust mechanism for everyone else's agents.
The $200M raised 11 months after $115M is not the story. The story is a $100M+ ARR company growing 60% with 30 seven-figure accounts. The capital is downstream of that GTM engine, not the cause of it.
1. LLM-as-distribution-channel (MCP, plugins) · The Wedge (Q1)This is the headline motion. Coralogix exposing Olly plus MCP and CLI interfaces, with half of enterprise customers already using them, is the clearest production-scale proof of this motion to date. When the buyer's agent finds and invokes your product directly, the structured surface is the product. This fits any company seed through Series D with an API-callable surface, and the first movers are defining the playbook right now.
2. Trust-layer-as-product · The Beacon (Q2)"Someone needs to watch the AI agents" is trust-layer-as-product stated as a category thesis. The compliance and observability artifacts are not documents that support the sale. They are the sale. This fits vendors selling into security-conscious buyers deploying autonomy they cannot yet audit, which by 2026 is most of the enterprise.
3. Land-and-expand seeding · Cross-Quadrant Amplifiers5,000 customers, 30 at $1M+, 60% revenue growth, and expansion from core logs into security and AI observability is the textbook expansion engine. The wedge was observability; the expansion is the agent era stacked on top of the same data foundation. Net expansion, not new logos, is carrying the growth, which is exactly what this motion produces when it works.
4. Anchor customer + reference flywheel · The Cathedral (Q4)IBM, Tradeweb, and JFrog are not decoration. They are peer-set proof that converts the next regulated buyer, including the financial institutions Coralogix is now landing through its India hub. Named, credible anchors in a category as conservative as observability do the institutional trust work that lets the product expand upmarket.
Four motions across three quadrants plus an amplifier is the signature of a company that has stopped running one play and started running a portfolio. That is the multi-motion era in a single funding announcement.
Across the 700+ enterprise AI transformations and 88 insurance AI vendor profiles we have mapped, the same pattern keeps showing up. Three layers worth naming:
Layer one. The buyer's interaction surface is collapsing into the agent. Engineers no longer want to learn your UI. They want to point their existing assistant at your data and ask a question. This is happening fastest in technical categories, but it moves into every category where the daily user is already working inside an LLM. The dashboard is becoming a fallback, not the front door.
Layer two. As the interface erodes, the vendor's differentiation migrates to the data layer and the trust layer. Coralogix kept its architecture and let agents query it, which is why "the architecture was already there" was the easy part. The hard part for most vendors is that they built for human eyes, not agent calls, and retrofitting structured invocability and provable safety is a rebuild, not a feature.
Layer three. Capital follows the company that has already made this shift, not the one promising to. A $1.6B valuation 11 months after the last round is the market pricing in a GTM engine that is already producing 30 seven-figure accounts and 60% growth. The raise is the consequence of the motion stack maturing. Founders who read this as "observability is hot, raise now" have the causality backwards.
The dashboard is becoming the least valuable thing observability vendors sell. The valuable things are the agent-callable data layer underneath and the trust apparatus that lets enterprises run autonomy they do not yet trust. Every infrastructure vendor is about to discover whether their product was built for human eyes or agent calls, and the answer determines whether they get to participate in the agent economy or get disintermediated by the agents themselves.
Three questions for founders and CROs:
By Q4 2026, every infrastructure and application vendor at Series B and above will be forced to ship an agent-callable interface (MCP or equivalent) as table stakes, and the ones who treat it as a roadmap item rather than a shipped product will start losing seven-figure renewals to competitors whose products their customers' agents can already reach.
What's your experience with the interface layer eroding in your own category? Drop a note or reach out directly.
.svg)
