Exciting news from Datadog AI Research: we trained a model to investigate production incidents, and it matches the best frontier models at a fraction of the cost.
Introducing Scout 1.0, the model powering Bits Investigation, our AI SRE.
On AlertEval, our benchmark of real
🎙️ We're joined by Yanbing Li, Chief Product Officer at @datadoghq, to talk about applying observability to non-deterministic AI agents, blurring the boundaries between software development and production workflows, and navigating emerging challenges in AI security and tokenomics.
stackoverflow.blog/2026/10/09/tak…
Another week, another round up - Shipped this week at @datadoghq 🚀
- Running a frontier model on every alert gets expensive fast. We launched Scout 1.0, our own model built specifically for incident investigation: matches frontier-model quality on real production alerts at 3 to 19x lower cost, and already powers Bits Investigation today. (@ThisIsOthmane)
- Agents built on @claudeai used to have limited production visibility. Claude Platform is now live with native OTel trace export, with Datadog as a launch partner. @claudeai@AnthropicAI
- Building test coverage meant scripting it and maintaining it by hand. Bits Testing is live now: it discovers critical user flows and generates tests for them, one click to adopt.
- Vulnerabilities that span several steps used to slip past pattern-matching scanners. AI-native SAST is live now, tracing how data actually moves through your code.
- Incident details live in Slack and bridge calls, not just telemetry? Bits AI Investigations already ingest the whole incident channel, not just monitor alerts.
- Signals explaining user drop-off used to be scattered across three products. Journey Monitoring is live now, unifying synthetic coverage, RUM, and product analytics in one view.
- A network issue meant context-switching across tools to find the blast radius? NDM now correlates issues to endpoints, network paths, and VPN tunnels on one screen.
Full write-up + docs → github.com/DataDog/shippe…
Your agents are in production. Now see everything they do.
Datadog Agent Observability is now available on the @claudeai Marketplace. Trace every step of your Claude-powered agents, correlated with the rest of your stack.
Learn more: bit.ly/3WtGX6Y
Two years ago, when we started building our SRE agent, we bet heavily on evals. We hand-labelled thousands of production incidents and built the infrastructure needed for our engineers to run this golden dataset every day as they iterated on the agent. That work underpins how we evaluate Scout 1.0. Specialized models start with knowing what good looks like in your domain.
Meet Scout 1.0, our first AI model built for investigating production incidents.
Scout matches frontier-model investigation quality at a fraction of the cost, which makes it practical to investigate more issues, earlier.
We trained Scout on realistic simulations built from
LAUNCHED: @claudeai now supports native OpenTelemetry trace export to @datadoghq Agent Observability.
Trace model calls, pinpoint failures, and inspect token usage and costs. No custom exporter required.
See how we debugged a failed flight booking in this demo 👇
LAUNCHED: @claudeai now supports native OpenTelemetry trace export to @datadoghq Agent Observability.
Trace model calls, pinpoint failures, and inspect token usage and costs. No custom exporter required.
See how we debugged a failed flight booking in this demo 👇
Meet Scout 1.0, our first AI model built for investigating production incidents.
Scout matches frontier-model investigation quality at a fraction of the cost, which makes it practical to investigate more issues, earlier.
We trained Scout on realistic simulations built from
You should be testing your agents in prod…
Offline evals and experiments still matter, but nothing beats the signal from A/B testing on real traffic.
Here’s how to do it in Datadog 👇
docs.datadoghq.com/experiments/gu…@datadogdevs@datadoghq
1K Followers 4K FollowingOperations & business growth leader. 20+ yrs turning around & scaling businesses. Built companies from zero. Now deep in AI systems & automated software.
879 Followers 461 Followingbuilding @withflintpay - commerce, not just payments, for developers and AI agents.
ex @Square
created: https://t.co/zLXtN0l6d7 https://t.co/JYyLIpFVEf
292 Followers 1K FollowingStaff Research Scientist (founding member of Datadog AI Research) | AI Agents for SRE/Observability | Agentic Post-Training | Observability Foundation Models
273 Followers 4K Following1988 graduate of West Texas A&M University with BBA in Computer Information Systems. Minor in business. Information Technology Consultant and Entrepreneur.