On-Call Policy

Purpose and Scope This policy defines how we handle urgent issues in production and staging (where staging may affect customers). As a services-based startup, we must minimize downtime, communicate clearly, and protect team wellbeing. This document sets the people on call, response timelines, communication rules, severity handling, post-incident documentation, and actions for non-compliance.


1) Coverage

  • Applies to every live project in production or staging where customer or client impact is possible.
  • Coverage is 24×7 for customer-impacting incidents.

2) On-Call Composition (per shift)

  • 1× Frontend Engineer
  • 1× Backend Engineer
  • 1× Team Lead (incident coordinator and decision maker)
  • Product Manager (PM) available during incidents to manage client communication.
  • If no Team Lead exists, the PM acts as TL as defined in this policy.

3) Scheduling

  • The PM prepares and maintains the weekly on-call schedule.
  • Each discipline has one assigned engineer per week, rotating fairly.
  • Swaps must be updated in the schedule and announced in Slack before the shift starts.

4) Alerts & Communication

  • Initial alert: via Slack and WhatsApp to on-call members.
  • All coordination happens in the project’s Slack channel.
  • PM drafts all customer messages, which are approved by the Team Lead (or acting TL).

5) Response Targets

  • Acknowledge alert within 10 minutes.
  • Start work within 15 minutes after acknowledgement.
  • Post concise, time-stamped updates in Slack until mitigation or resolution.

6) Severity Levels & Update Cadence

  • P0 (Critical): Full outage, data loss, payments down, or security issue.

    • Immediate all-hands by on-call trio + PM.
    • Updates every 10–20 min; client updates as needed.
  • P1 (High): Major degradation or partial outage.

    • Prompt engagement by on-call team.
    • Updates every 15–30 min; client updates as needed.
  • P2 / P3 (Moderate / Low): Minor impact or visual issues.

    • Acknowledge and confirm now.
    • Fix next working day.

7) Standard Incident Flow

  1. Alert received → acknowledge ≤10 minutes.
  2. Team Lead (or PM as TL) sets severity and directs response.
  3. Engineers implement fix, rollback, or feature-flag change; verify recovery.
  4. PM manages external comms after TL approval.
  5. PM documents the incident — timeline, impact, actions, root cause, and follow-ups.

8) Wellbeing & Compensation

  • If an incident requires after-hours work, the responder may arrive late the next day (coordinate with manager).
  • Exception: when the issue was caused by clear developer fault, no compensation applies.

9) Roles at a Glance

  • Team Lead (or PM as TL): sets severity, leads decision-making, approves customer comms.
  • Frontend/Backend Engineers: triage, mitigate, fix, verify, and update.
  • PM: owns schedule, manages clients, and writes the incident doc.

10) Failure to Comply & Disciplinary Action

Every on-call member is expected to meet the response standards and responsibilities in this policy. Failure to comply — including missed acknowledgments, delayed response, lack of communication, or negligence that extends customer impact — will lead to the following actions:

First Instance

  • Written warning from the Team Lead and PM.
  • No wellbeing compensation (late arrival, rest time, or any on-call benefit) applies for next coming incident.

Second Instance

  • Salary deduction calculated per hour for the time the problem remains unresolved or for the duration the client bears losses — whichever is longer.
  • The delay and loss will be logged in the incident record.

Third Instance

  • The engineer will be placed on a Performance Improvement Plan (PIP) for 1 month, during which:

    • Salary deductions may continue,
    • Work will be under closer review, and
    • Employment status will be formally reviewed after the PIP period.

Severe Negligence

Ignoring P0/P1 alerts, refusing to respond, or falsifying updates may result in immediate escalation to management and HR, and termination without prior warnings depending on severity.


Summary

  • Ack ≤10 min; start ≤15 min after ack.
  • On-call team: 1 FE + 1 BE + 1 TL (PM if no TL) + PM for client comms.
  • Slack for coordination; alerts via Slack + WhatsApp.
  • P0: update every 10–20 min  P1: 15–30 min  P2/P3: fix next day.
  • After-hours work → late arrival (unless fault).
  • Failure to comply → written warning → pay deduction → PIP + review.