A new report from Apica and Omdia shows that enterprises worldwide are facing a surge in telemetry data as agent-based AI moves from pilot projects to production, while many organizations say they lack the infrastructure needed to manage the development.
Apica today presents the results of a new report conducted by analytics firm Omdia that shows that companies worldwide are facing a sharp increase in telemetry data as agent-based AI moves from experimental projects to production environments.
The report, based on responses from over 300 IT decision-makers in North America and Western Europe, shows that many organizations do not yet have the infrastructure required to handle the data volumes that future AI agents are expected to generate.
According to Apica CEO Mathias Thomsen The development is greater than many organizations realize.
– Agent-based AI creates entirely new demands on how companies collect, process and analyze telemetry data. Organizations that do not build the right foundation today risk facing both increased costs and greater operational challenges when AI agents begin to be used on a larger scale, says Mathias Thomsen, CEO of Apica.
Telemetry data is already growing at a record pace
The report shows that 54 percent of organizations surveyed have seen their telemetry data volumes triple in the past 12 months, with an average growth of 3.7x year-over-year.
AI and machine learning workloads now account for around 43 percent of this increase, making AI the largest driver of growing data volumes in modern IT environments.
At the same time, 83 percent of respondents state that observability for AI is one of the top priorities in 2026.
Agent-based AI is expected to create a new data wave
The most notable conclusion in the report is the expected increase in telemetry data from agent-based AI systems.
Enterprises expect data volumes from these workloads to increase by an average of 9.5 times over the next two years, with 44 percent expecting growth of between six and a hundred times today's levels.
Despite this, nearly two-thirds of organizations say they are only “somewhat prepared” or less for the data growth that lies ahead.
More than one in five organizations have not yet analyzed the consequences that agent-based AI may have on the business's data management.
Infrastructure becomes a strategic management issue
Andi Mann, Chief Product & Technology Officer at Apica, believes that development requires a new way of thinking about observability and data management.
“The data shows what we’ve been telling business leaders since last spring. Agent-based AI is changing not only how we work with observability, but also the underlying architecture. A 9.5-fold increase in data volumes in two years is not a technical detail but a strategic infrastructure decision at the management level,” says Andi Mann.
The costs of observability are increasing rapidly
The report shows that the growing amount of telemetry data is already impacting companies' budgets.
Organizations report spending an average of $3.17 million per year on observability solutions, with nearly 20 percent spending more than $5 million annually.
At the same time, 81 percent are actively looking for alternatives that can reduce costs without reducing transparency in the business's systems and services.
Observability budgets are growing by an average of 28 percent per year. For 35 percent of organizations, the increase exceeds 52 percent annually.
AI projects are paused due to high monitoring costs
The survey shows that costs are already affecting companies' AI investments.
59 percent of organizations report that they have stopped or postponed at least one agent-based AI project because the costs of monitoring and observability have become too high.
The most commonly paused use cases are cybersecurity, compliance, and fraud prevention, areas where lack of monitoring can pose significant business risks.
– Scalability is the main reason why many organizations are unable to execute their agent-based AI projects. Without the right solutions, they expose themselves to operational, legal and security risks, says Torsten Volk, Principal Analyst at Omdia.
Telemetry pipelines are identified as a key factor
The report identifies telemetry pipelines as one of the most important factors in managing development.
54 percent of organizations have already implemented a telemetry pipeline and a full 97 percent have either implemented or are evaluating such solutions.
According to the report, companies that use telemetry pipelines are 50 percent more likely to be prepared for the data growth that agent-based AI is expected to generate over the next two years.
Among organizations that have already implemented agent-based AI at scale, they are 80 percent more likely to have avoided the cost challenges that otherwise slow development.
Apica sees a growing need for modern data infrastructure
Mathias Thomsen believes that the results show that AI strategies can no longer focus solely on models and computational capacity.
– Successful AI initiatives also require control over data flows, costs, and governance. Companies that invest in the right telemetry and observability architecture today will be better equipped for next-generation AI applications, says Mathias Thomsen.
An active investment window opens
The survey shows that the market is now in an active evaluation phase.
68 percent of organizations plan to evaluate changes to their observability platforms within the next six months.
Meanwhile, 70 percent plan to evaluate telemetry pipelines during the same time period.
According to the report, many companies are already in or about to enter a new investment cycle where the focus is on future AI infrastructure.
About the report
The study was conducted by Omdia on behalf of Apica and includes over 300 IT decision makers in North America and Western Europe. All respondents had direct budget responsibility or decision-making authority for investments in observability and telemetry data.
About Apica
Apica develops telemetry data management and observability solutions tailored for AI-driven operations. The company's platform is based on filtering, processing and governing telemetry data before it is sent to analytics platforms, aiming to reduce costs while improving data quality and governance.
Apica has been named a Visionary in the Gartner Magic Quadrant for Observability Platforms
Swedish
Danish
Norwegian
Finnish
English
German
Dutch