Cloudflare this week made Logpush Transformers generally available to all self-serve paying customers. At its core, the feature embeds an SQL engine into the log export pipeline: before logs leave the platform, you can use SQL to drop health-check noise, mask sensitive fields, and rename fields to your own schema.
Beyond the technical mechanics, what deserves our attention is the commercial implication: previously, enterprises had to deploy a separate log processing service — operating it, scaling it, and monitoring it themselves. Cloudflare has now folded that job into its own platform.
What this is
Logpush is Cloudflare's log export service, batch-pushing network, request, and security logs to downstream SIEM (security information and event management) platforms, data warehouses, or object storage. Transformers lets you write an SQL query that processes logs before they leave: drop fields, rename, redact, restructure — output stays NDJSON.
The critical point is "filter before export." Filtered records don't count toward downstream transfer or storage costs. Pull in 1TB of logs but ship only 200GB downstream — that's savings across two line items: egress bandwidth and downstream storage.
How the industry sees it
We've watched Cloudflare spend the past two years doing one thing consistently: folding capabilities that customers used to piece together from third parties into its own stack — zero trust, security services, and now log processing. Mid-tier cloud players like DigitalOcean and Akamai are also moving toward "integrated platforms" this year.
Supporters see this as the trend: enterprise IT is drowning in tools, and every consolidation helps. Opposition is also clear — writing SQL inside a cloud vendor's pipeline hands data-processing logic to the platform, creating fresh challenges for data sovereignty, auditability, and portability. One practitioner complained: "Cloudflare's SQL dialect differs from the standard — switching platforms means a rewrite."
Impact on regular people
For enterprise IT: Companies on Cloudflare can revisit their log pipelines, trimming operations and cloud spend; but they need to weigh the exit risk of having processing logic locked to the platform.
For working professionals: The "log cleaning" workload of data engineers and operations staff will be partly absorbed by cloud vendors. In the long term, these roles will need to migrate toward higher-value analysis work.
For consumer markets: Indirect impact — cheaper infrastructure gives SaaS and application services room to cut prices, but the pass-through is slow; ordinary users won't feel it in the short term.