BOSTON, Oct. 14, 2026 (GLOBE NEWSWIRE) — Coralogix, the data and AI platform for observability, today announced the launch of its Telemetry Lake Benchmarks, demonstrating how its infrastructure delivers fast query performance across petabytes of telemetry data stored in open formats within customers’ own cloud object storage.
The site provides continuously updated performance data across Coralogix environments, giving users visibility into query volumes, data ingestion and response times. As of October 14, 2026, the Coralogix platform had processed more than 4 billion queries and ingested approximately 50 petabytes of data over a seven-day period. The figures are refreshed every five minutes, and the underlying benchmark data is available through an API and downloadable Markdown file, allowing users to independently verify the results.
In addition to global performance data, the site provides benchmarks based on real production workloads, demonstrating query performance across different telemetry types and data volumes. In one benchmark account processing approximately 40 terabytes of data per day, half of queries, including complex data transformations, joins and window functions, returned results in approximately one second.
The Coralogix Telemetry Lake and the growing role of AI agents
The interactive site also provides a look under the hood at the architecture behind Coralogix’s Telemetry Lake, showing how telemetry is processed in-stream and stored in open formats within customers’ own cloud object storage. By maintaining operational and historical telemetry in a single, customer-owned data lake, Coralogix enables teams to quickly query their complete data history without the need for separate storage tiers or rehydration processes.
As AI agents take on a larger role in investigating incidents alongside human engineers, the ability to analyze vast amounts of historical telemetry without overwhelming an agent’s context window is becoming increasingly important. Through DataPrime and server-side aggregation, Coralogix enables agents to analyze hundreds of millions of data points while retrieving only a few hundred tokens of relevant findings, reducing the amount of information agents need to process.
Coralogix also recently tested how AI agents investigate incidents across five observability platforms using the same OpenTelemetry demo application and a $10,000 monthly budget per platform. Across ten scenarios, each repeated five times, Coralogix led in nine out of ten scenarios when evaluating agents’ ability to identify root causes and measure customer and financial impact. In one investigation, an agent identified 243 individual purchases that had bypassed fraud checks and determined their associated value, demonstrating how access to detailed telemetry can help agents move beyond diagnosing technical issues to understanding their business consequences.
“Seven years ago, we made the decision to store the full history of systems in an open format and in the customer’s cloud,” said Ariel Assaraf, Co-Founder and CEO of Coralogix. “At the time, engineers were the ones searching the information for answers when something went wrong. Today, AI agents are joining them, going through the same data to understand what happened and identify what caused the problem. When an agent receives only part of the picture, a single missing event can change the answer. That is why the ability to quickly access the entire history is becoming much more important.”
Coralogix’s Telemetry Lake is designed to ingest unsampled, full-fidelity telemetry while separating processing from storage. By analyzing data in-stream before it is stored, Coralogix enables faster alerting, real-time metric generation and greater control over how data is routed, helping customers reduce their total cost of ownership. Telemetry is then stored in open formats within customers’ own cloud object storage, where it remains accessible for future investigations without requiring a separate copy or rehydration process.
Explore the live benchmarks and Coralogix Telemetry Lake architecture at coralogix-benchmarks.com. Coralogix will publish new benchmarks here weekly.
To learn more about Coralogix, visit: https://benchmarks.coralogix.com/
About Coralogix
Coralogix is a leading provider of observability solutions that help engineering teams monitor, understand, and act on their system data in real time. Through Olly, its built-in AI investigator, along with MCP and CLI interfaces for agent-based workflows, Coralogix supports three modes of operation on a single data foundation: human-led investigation, conversational AI collaboration, and fully automated agent workflows. Built on a streaming engine and a schema-free data lake with customer-owned storage in open formats, Coralogix captures all production data, not a sampled subset, and makes it available to human engineers and AI agents alike. The company is trusted by thousands of teams worldwide across financial services, media and entertainment, retail, and cloud-native enterprises.
Media Contact
coralogix-pr@walkersands.com
Media gallery
