Daily morning health roll-call
Every morning Zaira checks latency, error rate, key monitors, and host counts and posts a one-line verdict — “all green” or “these 3 monitors are over their targets” — to your Slack channel.
Datadog AI Agent
Connect your Datadog to Zaira autonomous AI agent. Describe what you need in plain language, Zaira will use Datadog actions like log search to get the job done for you.
01Capabilities
What can your Datadog AI agent do?
The full Datadog observability stack in one conversation — dashboards, monitors, logs, metrics, and APM tracing — no code, read-only. Zaira carries out tasks using the supported actions listed here.
- Search and query logs across your infrastructure
- Query metrics and analyze timeseries data
- Search traces and inspect APM span performance
- Retrieve dashboards and notebook contents
- View monitors, monitor groups, and their status
- Search and inspect incidents
- List hosts, services, and their health states
- Search events for org-wide visibility
02Use Cases
Automate your Datadog tasks with Zaira effortlessly
See how people are automating their Datadog tasks with Zaira while they sleep.
Weekly reliability & monitor digest
Each week Zaira reads your monitors, incidents, and events and files a plain-English report for your team: what triggered, what resolved, what broke, and service health.
Incident roundup & follow-ups
When an incident resolves, Zaira summarizes what happened, the impact, and what's still open — delivered to Slack, email, or a postmortem Doc before anyone has to write it.
Error & log spike alerts
Zaira compares error logs against the same window yesterday and flags unusual spikes in a service — with the top messages — before they become an outage.
Trace & APM investigation briefs
Why did checkout time out?” — Zaira walks the traces and spans, then attaches a plain-English explanation of where the slowness came from to the incident.
03How it works
How to set up your Datadog AI agent
Connect Datadog in one click, put monitoring on autopilot, and watch health roll-calls and alerts delivered to Slack and Docs — no code. Run tasks on demand; recurring runs depend on the workflow's supported capabilities.
01
Connect your Datadog in one click
Sign in with your Datadog account (OAuth) and Zaira wires up the official Datadog MCP server for you — no code, no config files, nothing to build.
02
Put your monitoring on autopilot
Tell Zaira what to watch and when — daily health roll-calls and weekly reliability reports to Slack or email, incident roundups, and error-spike alerts on a schedule.
03
Nothing breaks silently anymore
Daily health roll-calls and weekly reliability reports land in Slack or email before standup — with error spikes and incident roundups surfaced on schedule. Ask “is the app healthy?” anytime. Outages get caught before anyone notices.
05FAQ
Frequently asked questions
See the answers to the questions people ask about how Datadog works with Zaira.
You choose a task and Zaira uses supported Datadog actions, including log search, to carry it out. Tasks can run on demand; recurring or event-triggered runs depend on the workflow's available capabilities.
No. The Datadog AI integration from Zaira works with zero code. You sign in with your Datadog account in one click (OAuth) and Zaira wires up the official Datadog MCP server for you — no code, no config files, nothing to build. Your Datadog admin may need to grant the MCP Read permission to your role first, a one-time setting. After that you just ask in plain language.
Zaira can read your logs, metrics, traces, dashboards, monitors, incidents, and host health — and act on them — posting daily health roll-calls, weekly reliability reports, incident roundups, and error-spike alerts to your Slack, email, Docs, or Sheets.
Yes. Tell Zaira what to watch and when — a morning health roll-call, a weekly reliability digest, an instant alert whenever error logs spike. Both run automatically, nothing to build or maintain.
Bits AI lives inside Datadog (or its incident Slack channels) and is built for SRE-style investigation, while Scheduled Reports email recurring PDFs of dashboards. Zaira reads the same data and delivers plain-English summaries to the whole team wherever they work — Slack, email, Docs, and Sheets — without any dashboard-watching or query building.
Datadog's official MCP server is on-demand: you open a chat, ask a question, and it answers — but nothing watches your systems, nothing runs on a schedule, and setup involves role permissions and site endpoints. Zaira connects once (the same official server, wired for you), then adds the watching, scheduling, and delivery that MCP doesn't have.
The integration is read-focused — metrics, logs, traces, monitors, incidents, and hosts. Zaira reads, summarizes, and reports; actions like creating monitors or muting hosts stay in Datadog.