Today, Chronosphere, which makes an observability platform that is focused on microservices and containers, is announcing the launch of AI-Guided Troubleshooting capabilities, a major advancement that redefines how engineering teams investigate and resolve production incidents. The new set of capabilities combines AI-driven insights with deep environmental context through a Temporal Knowledge Graph. With this context, Chronosphere delivers highly accurate root-cause insights that enable engineers to resolve issues faster and with greater confidence.
“Chronosphere is a 2022 startup,” said Martin Mao, CEO and Co-founder of Chronosphere. “I ran observability at Uber, and saw the challenges they faced as they grew. The volume of data grew, as did the number of customers. The tools they used were built for more applications that were more legacy, so they were expensive and not the right solution. In 2019 I left Uber and founded this company, which was enterprise, and we built a SaaS solution on top of it.”
Mao said that the competition in their space was limited.
“There is really just two or three public companies left, because of the transition to containerized environments,” he indicated.
Mao indicated that Chronosphere differentiated itself from the market in a couple of ways.
“The first is that we used a disruptive pricing model,” Mao stated. He emphasized that the model is unique, and different from others out there.
“We assess how useful it is, analyzing the income stream and raw data – and we only charge for that,” he said. Instead of pricing based on host or VM, Chronosphere only charges for the useful data that you choose to retain in the Chronosphere Observability Platform. With our Control Plane, you can analyze and shape your observability data to fulfill your existing dashboard and alerting needs without having to store all the data in the raw form. Unlike other observability solutions that require you to store, query, and pay for all data, whether or not it is useful, Chronosphere pre-computes the data to get rid of the waste and efficiently delivers it to the glass.
Every organization’s needs are very different, so Chronoscope wants to work with the customer to drive the right value at the right price, without the fear of overages or huge price hikes when their data complexity inevitably grows. For the Chronosphere Telemetry Pipeline, they follow a straightforward pricing model, based solely on data throughput – the volume of raw data you transmit through the pipeline.
The second differentiation relates to the quality of Chronoscope’s features themselves.
“We have better analytics and features on top of the raw data itself, which makes it much faster,” Mao stated. “We are aiming for the top end of the market.”
The new AI-Guided Troubleshooting capabilities are a major advancement that redefines how engineering teams investigate and resolve production incidents.
“As the amount of AI-generated code is increasing, observability systems detect and help to remediate this, using AI powers,” Mao said, “It makes humans far more efficient, and is the only viable way to sustain a large amount of code. The trick to having it work is to have as complete a knowledge graph as possible. It makes agents more effective. This is necessary because most availability data is not normalized, and not standard. We can scrub all the raw data.”
Mao indicated that the knowledge graph was applications that talk to each other.
“It tells you things like error rates and latency, because in observability, you want to see a system before it went wrong,” he said.
Research from MIT and the University of Pennsylvania found that generative AI spurred a 13.5% increase in weekly code commits, signifying a surge in code velocity and change volume. Despite these advancements in software development, troubleshooting remains primarily manual and relies heavily on intuition, resulting in slower mean time to resolution (MTTR) and greater on-call stress.
Chronosphere’s AI-Guided Troubleshooting capabilities close this gap by combining AI reasoning with the Temporal Knowledge Graph – that living, queryable map of an organization’s services, infrastructure, and their relationships. It accounts for system changes and even human input. Unlike observability tools that run on proprietary or standard data inputs, it also integrates custom application telemetry, providing the deep context needed for effective root-cause analysis.
“This is a net-new set of features, providing additional capabilities on data we already had,” Mao said. “One difference is that it’s automated, and you no longer have to manually sift through information on the site. Another differentiation is the way in which it integrates custom application telemetry.”
Chronosphere’s AI-Guided Troubleshooting introduces four core capabilities. The first is Suggestions, proactive, plain-language insights that guide investigations toward likely causes, and which are backed by data, not guesswork. The second capability is the Temporal Knowledge Graph, a continuously updated map of services, dependencies, and custom telemetry, capturing full system context. The third new feature is Investigation Notebooks, which are persistent workspaces that document every step, piece of evidence, and conclusion, turning investigations into reusable institutional knowledge. Finally, the fourth new capability is Natural Language Assistance, in which engineers can now build queries and dashboards using natural language, accelerating data exploration.
“For AI to be effective in observability, it needs more than pattern recognition and summarization,” Mao said. “Chronosphere has spent years building the data foundation and analytical depth needed for AI to actually help engineers. With our Temporal Knowledge Graph and advanced analytics capabilities, we’re giving AI the understanding it needs to make observability truly intelligent – and giving engineers the confidence to trust its guidance.”
With this context in place, the system then applies Chronosphere’s advanced analytics to surface the most meaningful next steps in an investigation. At each stage, it explains what’s been analyzed or ruled out, allowing engineers to stay in control while AI accelerates every phase of the troubleshooting process. As engineers zero in on a root cause, investigations are fed into the Temporal Knowledge Graph, so future suggestions get smarter.
“The uniqueness of our solution looks almost boring, with all the data scrubbing,” Mao said.
In addition to AI-Guided Troubleshooting, Chronosphere announced the general availability of its Model Context Protocol (MCP) Server.
“Two things are unique about his MCP server,” Mao said. “We have the largest foundational models in the world, which give us a bleeding edge view. We are also attempting to figure out an industry standard way to do this – standardized for the industry.”
AI-Guided Troubleshooting, including Suggestions and Investigation Notebooks, is in limited availability today, with full general availability planned for 2026. MCP integration is available now for all Chronosphere customers.