Skip to main content

​
Overview

WHAWIT sits on top of your existing observability tools to interpret telemetry, reduce MTTR, and recover engineering time from manual incident triage. In this quickstart, you will connect WHAWIT to your stack, configure your first services, and walk through an example incident in the On-Call Hub.
This guide assumes you already have at least one observability platform in place (for example Datadog, New Relic, CloudWatch, or Elastic) and that it is receiving production telemetry.

​
Step 1: Access WHAWIT

1

Sign in or create an account

  1. Go to https://app.whawit.ai.
  2. Sign in with your existing account or create a new one.
You should see the WHAWIT dashboard with no services configured yet.
2

Invite your team (optional)

Add SREs, platform engineers, and developers who participate in incident response so they can access the On-Call Hub and incident history.

​
Step 2: Connect observability sources

  1. In WHAWIT, navigate to Integrations → Datadog.
  2. Follow the prompts to create an API key in Datadog with read-only access.
  3. Paste the API key into WHAWIT and select the services or tags you want WHAWIT to monitor first.
Start with a small set of critical services so you can validate WHAWIT’s behavior before rolling out broadly.
  1. Open Integrations and choose your provider (CloudWatch, New Relic, Elastic, or others).
  2. Provide the required credentials (for example, AWS IAM role, New Relic API key, or Elastic endpoint and token).
  3. Confirm which log groups, metrics, or indexes WHAWIT should analyze.
WHAWIT does not replace your observability stack. It reads telemetry from your existing tools and adds an intelligence layer on top.

​
Step 3: Define services and environments

  1. Go to Settings → Services.
  2. Create a service for each critical application or microservice (for example, checkout-api, search-service).
  3. Map each WHAWIT service to the relevant telemetry from your observability tools (for example, Datadog tags, AWS log groups).
When a service is correctly mapped, you should see recent telemetry samples in the WHAWIT UI.
Define environments such as production, staging, or sandbox so WHAWIT can prioritize incidents and apply different alerting rules.

​
Step 4: Experience your first intelligent incident

1

Trigger or wait for an incident

  • If you have an upcoming test window or game day, introduce a controlled failure in a non-production environment.
  • Otherwise, wait for WHAWIT to detect an anomaly in your production telemetry.
2

Use the On-Call Hub

  1. Open the incident in the On-Call Hub.
  2. Read the natural-language summary that WHAWIT generates from logs, metrics, and traces.
  3. Review the event timeline to see what changed, in what order, and which services were impacted first.
  4. Use the deep links into your existing tools if you need to validate details.
The goal is to reach a credible hypothesis about the incident cause in minutes, not in the first 30–60 minutes of the outage window (full discussion in the WHAWIT extended version).
3

Capture learnings and improvements

After resolution, use WHAWIT’s recommendations to:
  • Identify which code or configuration changes were involved.
  • Capture post-incident notes directly in the incident record.
  • Optionally, export insights into your ticketing system or backlog.

​
Next steps

Now that you have WHAWIT connected and have seen an example incident, explore these next steps:

Deep dive into architecture

Understand how WHAWIT’s intelligence layer models your systems and telemetry.

Executive overview

Share a short, ROI-focused overview of WHAWIT with stakeholders.

Watch demos

See Arcade walkthroughs of WHAWIT handling incidents and telemetry overload.

Read the whitepapers

Dive into the research on downtime economics and autonomous reliability engineering.