The challenge
Ten thousand hotels running their physical operations on one platform ties Mews engineering to continuous, zero-downtime global commerce. At that level, visibility and production truth is completely essential, not only for the health of the infrastructure, but the health of the business.
By early 2025, the telemetry architecture had not kept pace. Raw data ingestion had grown 63% against monolithic vendor pricing, pushing observability costs past $500,000 a year on 130TB a month.
Almost none of it was usable. Legacy routing sent critical telemetry through unindexed debug logs, so metrics and distributed tracing were functionally absent. Mews was paying hyperscale prices for data it could not query.
We were paying a premium for hyperscale data ingestion while yielding very few actionable insights.
Terry Brown, Sr. Director of Engineering, Mews
Choosing Coralogix
Closing that gap needed a different relationship with the data, not a tool swap. Traditional monitoring vendors run a transactional model that penalises data growth. The more a company sees, the more it pays, and the vendor has no reason to help it see more for less.
We needed an integrated engineering partner willing to co-author our migration roadmap, not a vendor auditing our data volume.
Terry Brown, Sr. Director of Engineering, Mews
Mews executed the cutover against a hard target date to remove dual-running overhead, and maintain operational visibility. This ran successfully, and that is when the value began.
The outcome: from SRE tool to company-wide competency
Coralogix decentralised platform ownership at Mews. Real-time metrics, distributed tracing and SLO error budgets went into the hands of 850 cross-functional users.
With Coralogix, Mews can:
- Give every team its own metrics, traces and SLO error budgets, so the team that deployed the code owns the operational truth of that deployment
- Run an Observability Champions academy spanning engineering, customer support and marketing
- Add users without adding cost, because Coralogix decouples ingest from access and charges no per-seat fees
- Query the whole estate in DataPrime and follow requests through distributed microservices in APM
Supporting agentic engineering
How code gets written at Mews has changed with agentic coding. Mews has become AI-native, and the Coralogix CLI and MCP server put telemetry in the terminal alongside the agents doing the work. Mews built infrastructure to route their requests to the best target, meaning engineers are burning tokens in the most efficient possible way.
We went AI-native deliberately, and our engineers expect production truth to be in the terminal with them. Coralogix is how we give them that.
Terry Brown, Sr. Director of Engineering, Mews
The results
- 850 active users across engineering, customer support and marketing, on a platform only SREs had touched
- Metrics and distributed tracing in production for the first time, giving unprecedented visibility, insight, and compounding value
- A single-pass cutover against a hard target date, with no dual-running period
- AI-native observability unlocked, bringing production truth into the terminal, where engineers can use it without context switching
Coralogix turned observability at Mews from a siloed SRE tool into a shared engineering competency, owned by every team that ships, powering everything from production releases to operational decision making.