npx @buildinternet/releases get datadog-blogDatadog introduced Scout 1.0, its first AI model specialized for investigating production issues, which powers its AI SRE Bits Investigation. Trained on GLM-5.3 Flash with reinforcement learning over historical incidents and telemetry simulations, Scout 1.0 medium matches GPT-6.1 Sol high and Claude Opus 5.5 high at 3-8x lower estimated inference cost, with the xhigh mode matching Fable xhigh at 4x lower cost.
Operations Monitoring, included with RUM at no additional cost, lets teams define application-controlled steps like "process payment" with their own starts, outcomes, and durations, then track availability and latency independently using out-of-the-box metrics rum.measure.operation and rum.measure.operation.duration. Teams can set separate availability and latency SLOs per operation and use Bits AI to investigate degradations.
Datadog examines how the legacy device plugin model limits GPU utilization through fixed MIG partition profiles, opaque GPU counts that can't express memory or topology requirements, and a lack of scheduler-managed capacity accounting. The post covers the three mechanisms Kubernetes Dynamic Resource Allocation uses to raise utilization and how Datadog GPU Monitoring can track the resulting cost savings.
Datadog describes building a service criticality system that combines incident postmortem history, APM/RUM error traces, traffic volume, and on-call monitors into high, low, or unknown levels with a confidence score. The postmortem signal used an LLM to extract involved services from several hundred SEV-1 and SEV-2 postmortems, validated against Software Catalog.
Cross-cluster queries let teams and AI agents search selected BYOC clusters with a single query in the Log Explorer, with Datadog running the query against each cluster and combining matching results while logs stay in their original locations. The same scope works in dashboards, log monitors, and the MCP Server via search_datadog_logs.
Datadog Incident Response is now covered by Datadog for Government's FedRAMP High certification, bringing paging, incident coordination, automation, and postmortem workflows into the GovCloud US1-FED environment. Datadog says it is the only incident response platform with FedRAMP High certification at publication time.
Datadog Action Management's Visual Labeler lets teams navigate their web app via the test recorder Chrome extension, select an element, and name the interaction without writing code or deploying, including on authenticated and gated pages. Labels apply retroactively to matching interactions already recorded within the 15-month retention window, and labeled actions can be reused across Product Analytics funnel steps, dashboards, and Notebooks.
Datadog CI/CD Optimization brings pipeline and test data together with commits, logs, runners, and infrastructure to cut pipeline duration, surface flaky tests, and skip tests unaffected by a change. It includes CI/CD and test monitors, Flaky Management, Auto Test Retries, Early Flake Detection, PR Gates, and Test Impact Analysis.
Observability Pipelines Packs adds seven preconfigured, source-specific Packs for Cisco ASA, Fortinet FortiGate, Palo Alto, CrowdStrike FDR, SentinelOne Cloud Funnel, Windows, and Zscaler that drop, dedupe, and sample low-value security logs before they reach Exabeam while leaving raw payloads intact. Destination-agnostic Packs for Active Directory, Okta, Microsoft 365, and Abnormal.ai can be paired with an Exabeam destination for identity and collaboration coverage.
Guide explains collecting and querying Databricks telemetry through system tables, covering compute, job, query, data quality, and Model Serving performance, plus complementary monitoring via job notifications, the Jobs API, and the Spark UI.
New guide covers using the Databricks integration to monitor jobs and clusters, catch data quality issues with Data Observability, monitor Model Serving endpoints, track DBU costs with Cloud Cost Management, and contextualize telemetry with Reference Tables. The integration collects Spark metrics and logs via the Agent, job run data via API polling, and cost and lineage data via system table queries.
Datadog Code Security now manages Supply Chain Firewall as an organization-wide control, letting teams define allowlists and blocklists once in Datadog and apply them across workstation and CI installations, with firewall events reporting to a single feed. A new Supply Chain Firewall GitHub action brings the same protection to CI workflows, and Code Security retroactively scans recorded installations against updated malicious-package intelligence to alert on packages identified as malicious after install.
Datadog extended Apache DataFusion to execute a single query across multiple machines, releasing the framework as open source on GitHub and crates.io under Apache 2.0. The post reports up to 10x faster completion on heavy queries and benchmarks against Ballista, Spark, and Trino on TPC-H and TPC-DS.
Datadog describes an async-aware profiling model that tracks parent-child relationships between asyncio tasks to build "stacked stacks," preserving the logical execution path that traditional flame graphs flatten. Making the model practical in production also cut profiler overhead by more than 60% while producing higher-fidelity profiles.
Datadog now supports Real User Monitoring and Product Analytics on Shopify checkout and Salesforce Experience Cloud using each platform's supported extension points: a Liquid snippet plus a custom pixel for Shopify, and a Salesforce-specific bundle loaded as a static resource via loadScript for Lightning Web Security. Core Web Vitals and Session Replay remain storefront-only on Shopify, and unhandled promise rejections aren't captured under LWS.
New guide covers building a Jev rubric that scores all criteria in a single request, wired into both online evals on live spans and offline evals in Datadog experiments.
RUM Remote Configuration lets teams change supported SDK settings — sampling rates like rum.sessionReplaySampleRate and rum.traceSampleRate, privacy settings, event tracking, and app attributes — from the Datadog UI without redeploying frontend code. Existing applications add the remote configuration ID to their initialization code and enable it; published settings override SDK values on next initialization, with separate view and edit permissions and an audit log of last modification.
Tap to Parse is an AI-assisted capability that analyzes sample logs to identify their stable structure, distinguish fields from variable values, and generate a parser you can review before applying it. It is generally available in the Log Explorer and Log Pipelines, and available upon request in Observability Pipelines.
Datadog fine-tuned Qwen3.5-9B on investigation traces from GLM-5.3, reaching 87% of the teacher model's recall at $0.003 per investigation versus $0.06 in API charges, a roughly 20x cost decrease. A single 40 GB A100 supports about 100,000 investigations per week.
Code Execution is now generally available in the Datadog MCP Server, letting agents run sandboxed JavaScript that queries multiple Datadog APIs through the execute_code and search_datadog_sdk tools, branch and join results in code, and return only the evidence needed. Authentication stays outside the sandbox, with each API operation typed and subject to the caller's permissions.
