On-call that pages you
with the answer attached.
WHAWIT includes full on-call management — schedules, rotations, overrides and escalation policies. You do not need PagerDuty or Opsgenie alongside it.
The difference is what arrives with the page. A standing team of AI agents has already worked the incident — reasoning about what to check next rather than following a runbook you had to write first — so the notification on your phone carries the root-cause hypothesis, the evidence and a suggested fix.
From alert to acknowledgement
The page arrives already investigated
Traditional on-call tools tell you something broke. WHAWIT tells you what broke, why it probably broke, and what to try — in the same notification.
A signal becomes an incident
Alerts arriving from your providers are correlated and deduplicated, so one failure pages once instead of forty times.
Radar investigates first
Before a human is involved, the AI agent pulls the logs, metrics and recent deploys around the failure and forms a root-cause hypothesis with the evidence attached.
The schedule resolves who is on call
WHAWIT evaluates the rotation in its own timezone — layers, handoff times, restricted hours and any active override — and returns the engineer who is actually responsible right now.
You get paged with the triage in hand
The notification goes out on WhatsApp, SMS, voice call, Slack, Teams, Discord or email, and it carries the summary, the severity, what WHAWIT already found and a one-tap link to acknowledge.
No acknowledgement, next level
Each escalation level has its own timeout. If nobody acknowledges in time, WHAWIT moves to the next level on its own, and can repeat the whole chain until someone picks it up.
An on-call team that never sleeps,
and never needed your runbook
WHAWIT does not run a decision tree you had to write first. A standing team of five AI agents works the incident the way people would — deciding what to look at next based on what they just found, not on a rule somebody anticipated months ago.
Incident Commander
Decides what the team is dealing with, opens and updates the incident, and calls the escalation.
Log Analyst
Queries the logs, groups the errors, and separates the signal from the thousand lines around it.
Infra Specialist
Correlates the failure against infrastructure, recent deploys and the code in your repository.
Comms Lead
Notifies whoever needs to know, on the channel that reaches them, with what has been established so far.
Knowledge Engineer
Writes back what the team learned, so the next occurrence starts from what this one cost.
What they can reach for
No rules to pre-write
There is no runbook to author, no alert-to-action mapping to maintain, no if-this-then-that to keep current as the system changes. The agents reason about what they find.
It runs whether or not anything fired
The team works on a cycle, triggers immediately on an alert, and schedules its own follow-ups. The interval tightens on its own while an incident is live.
It gets better at your systems
What each incident establishes is written back to the knowledge base, so the next occurrence is met with what the last one cost you to learn.
What is included
A complete on-call system, not an integration
Everything below is built into WHAWIT and configured from the same place you manage incidents.
Who is on call, resolved correctly at 3 AM
A schedule lives in its own IANA timezone and is built from layers — Primary, Secondary, as many as you need. Each layer rotates through its members daily, weekly, or on a custom shift length you set in hours, with an explicit handoff time.
- Daily, weekly or custom shift lengths
- Multiple layers per schedule, each rotating independently
- Restrictions so a layer only covers business hours or weekends
- Overrides for time-off and swaps, on one layer or all of them
- Timezone-correct handoffs, including shifts that cross midnight
- Look up who is on call now, or the shifts for any future window
Escalation that keeps going when nobody answers
A policy is an ordered chain of levels. Each level names its targets and how long it waits for an acknowledgement before moving on. Miss the whole chain and the policy can repeat it.
- Per-level acknowledgement timeout in minutes
- Target an individual, a whole schedule, or an escalation team
- Per-level channel overrides — start on Slack, end on SMS
- Repeat the entire chain N times if the incident is still unacknowledged
- An organization-wide default policy as the fallback
- Preview a policy to see exactly who it would page, before an incident
Reach people where they actually look
Every escalation level delivers on the channels you choose. The message carries the severity, the summary, what the investigation found and a link to acknowledge in one tap.
- WhatsApp, SMS and voice calls for the middle of the night
- Slack, Microsoft Teams and Discord for the working day
- Email as the always-there fallback
- Acknowledge from the link to stop the escalation immediately
How it compares
Everything they page with, plus the investigation
WHAWIT covers the on-call surface teams expect from PagerDuty or Opsgenie. What it adds is the work that normally starts after the phone rings.
| Capability | WHAWIT | PagerDuty / Opsgenie |
|---|---|---|
| On-call schedules, layers and rotations | ||
| Escalation policies with acknowledgement timeouts | ||
| Overrides for time-off and shift swaps | ||
| Alert correlation and deduplication | ||
| WhatsApp as a first-class paging channel | ||
| Autonomous agent team that investigates without pre-written runbooks or rules | ||
| Investigation continues on its own cycle, whether or not anyone is watching | ||
| Root-cause investigation across your logs, metrics and code before paging | ||
| Root-cause hypothesis and suggested fix inside the page itself | ||
| Reads your existing observability stack read-only | ||
| Postmortem drafted from the incident timeline |
Comparison reflects the on-call and incident-response capabilities described on each vendor's public documentation as of July 2026.
Already on PagerDuty or Opsgenie
Nobody has to migrate on day one
WHAWIT ingests alerts from the paging tool you already run, so you can evaluate it against live incidents before moving a single rotation.
Keep paging where you are
Point PagerDuty or Opsgenie at WHAWIT. Your existing rotations keep paging exactly as they do today — nothing changes for the people carrying the phone.
Watch it investigate the same incidents
WHAWIT picks up those alerts, correlates them, and runs its investigation in parallel. You get to compare what it found against what your team found, on real incidents, without betting anything on it.
Move the schedules over
When the investigations are earning their keep, rebuild the rotations in WHAWIT and turn the old escalation off. One system, one bill, and the page now arrives with the answer.
Questions
On-call, answered directly
Yes. WHAWIT handles on-call natively. It includes on-call schedules, rotating shifts, multiple rotation layers, overrides and escalation policies as built-in features, and it pages the responder on WhatsApp, SMS, voice call, Slack, Microsoft Teams, Discord or email. On-call is a core part of the product, not an integration with a third-party paging tool.
Yes. You create on-call schedules directly in WHAWIT. Each schedule has its own IANA timezone and is composed of layers, such as Primary and Secondary, that rotate through their members daily, weekly, or on a custom shift length measured in hours. Each layer has an explicit handoff time, and you can restrict a layer to specific days and hours so it only covers business hours or weekends.
No. WHAWIT does not require PagerDuty, Opsgenie or any other paging tool. Its on-call scheduling and escalation are built in, so WHAWIT can be the only on-call system your team runs.
Yes. WHAWIT covers the on-call surface teams rely on in PagerDuty and Opsgenie — schedules, rotations, overrides, escalation policies with acknowledgement timeouts, and multi-channel paging. It adds a step those tools do not have: an AI agent investigates the incident across your logs, metrics and code before the page goes out, so the notification arrives with a root-cause hypothesis and a suggested fix already attached.
No. WHAWIT's on-call is agentic rather than rule-based. There is no decision tree to author and no alert-to-action mapping to maintain: a team of AI agents reasons about each incident from what it finds, deciding what to investigate next based on the previous step rather than following a path someone anticipated in advance. You configure who is on call and how escalation should reach them; you do not have to encode how to investigate.
WHAWIT runs a standing team of five specialist AI agents — an Incident Commander, a Log Analyst, an Infrastructure Specialist, a Communications Lead and a Knowledge Engineer. Each has its own tools: querying logs, grouping errors, searching your repository, searching the knowledge base, analyzing context, opening and updating incidents, notifying responders and attempting recovery. They work on a cycle, trigger immediately on an alert, and schedule their own follow-ups, so the investigation continues whether or not a human is watching.
Yes. The agent team runs continuously on its own cycle rather than only reacting to a page. It triggers immediately when an alert fires, schedules its own follow-up work, and tightens its cycle while an incident is live. What each incident establishes is written back to the knowledge base, so recurring failures are met with what was learned the last time.
An escalation policy is an ordered chain of levels. Each level names its targets — an individual, an entire on-call schedule, or an escalation team — and sets how many minutes it waits for an acknowledgement. If no one acknowledges within that window, WHAWIT automatically escalates to the next level, and the policy can repeat the whole chain a configurable number of times if the incident is still unacknowledged.
WHAWIT pages on-call responders over WhatsApp, SMS, voice call, Slack, Microsoft Teams, Discord and email. Channels are configurable per escalation level, so an incident can start in a Slack channel during working hours and move to SMS and WhatsApp if it is still unacknowledged.
The page carries the incident severity and summary, what the AI investigation already found — the root-cause hypothesis, the supporting evidence and a suggested fix — and a link to acknowledge the incident in one tap. Acknowledging from that link stops the escalation immediately. The intent is that the responder starts from a finished investigation rather than an empty dashboard.
Yes. WHAWIT can ingest alerts from PagerDuty and Opsgenie so you can adopt it without switching everything at once, and enrich the incidents those tools raise. Most teams use this as a migration path: keep the existing paging in place, watch WHAWIT investigate the same incidents, then move the schedules over once they trust it.
Yes. An override replaces whoever the rotation would otherwise select for a given time window. You can scope an override to a single layer or apply it across every layer in the schedule, and attach a note explaining why it exists.
Put your rotation on it
Bring one schedule and one escalation policy. The next time it fires, see what the page looks like when the investigation is already done.
Questions about rotations or a migration? [email protected]. Already running WHAWIT on-call? [email protected].