Datadog Adds Bits AI Tools for Building, Evaluating and Monitoring AI Agents

Datadog announced more than 100 new capabilities at DASH 2026, including additions to its Bits AI product line aimed at helping teams build, test, secure and monitor AI agents.

Bits AI expands into agent development and operations

According to Datadog’s press release, the company introduced new Bits AI capabilities including Agent Evals, Bits Agent Builder and Agent Console. Datadog says these tools are intended to support the creation, debugging, securing and monitoring of AI agents.

Datadog’s AI-focused DASH 2026 roundup describes Bits Agent Builder as part of a broader set of AI workflow announcements. The company says the builder is designed to help teams create agents that can interact with operational data and workflows. Datadog also names Agent Console as a tool for managing and observing agents once they are in use.

Agent Evals, as described in Datadog’s announcement, is positioned around evaluating agent behavior. That reflects a broader operational challenge for companies adopting AI agents: teams need ways to test whether agents are producing reliable outputs, taking appropriate actions and staying within intended boundaries. Datadog’s materials frame these new features as part of the company’s observability and security platform rather than as standalone AI development tools.

Monitoring AI coding tools

Datadog’s press release also says the new Bits AI capabilities include support for monitoring coding-agent tools such as Claude Code, Cursor and GitHub Copilot. The company describes this as a way to help organizations understand and manage AI-assisted software development activity.

The announcement does not provide independent performance benchmarks for these features. Instead, Datadog’s framing focuses on the operational issues that can arise as software teams adopt AI tools across development, security and infrastructure work.

Datadog’s DASH 2026 keynote roundup says the company is adding Bits AI support for developer and operations workflows, autonomous detection and remediation, and AI-driven release validation. Those additions fit with Datadog’s existing focus on software reliability, monitoring and incident response.

Autonomous investigations and remediation

In its AI-focused DASH 2026 roundup, Datadog says Bits AI includes autonomous investigations and integrations based on the Model Context Protocol, commonly known as MCP. Datadog presents these integrations as a way for Bits AI to connect with external systems and operational context.

The company’s DASH materials also describe autonomous detection and remediation features. Datadog says these are intended to help teams identify issues and take corrective actions more quickly. The source documents do not say that human oversight is removed, and Datadog’s announcements frame the features as tools to support engineering and operations teams.

Security and complexity are central to the pitch

Datadog’s press release says the 100-plus capabilities are meant to help customers manage growing AI and security complexity. The company ties the Bits AI announcements to a broader set of platform updates across observability, security and software delivery.

For organizations adopting AI agents, the practical significance of Datadog’s announcement is that agent behavior is being treated as an operational surface that needs monitoring, evaluation and controls. Datadog’s new products target that problem from several angles: creating agents, evaluating them, observing them in production and connecting them to incident and developer workflows.

Because the information comes from Datadog’s own press release and company blog posts, the claims should be read as product announcements rather than independent assessments of effectiveness. The announced features nevertheless show how observability vendors are adapting their platforms for teams that are adding AI agents and AI-assisted coding tools to everyday software work.