Service Level Agreement (SLA)

Overview

In the Citizen Complaint Resolution System, the SLA or Service Level Agreement specifies how long a grievance can remain in a specific status before it is automatically escalated to the next level (for example, assigned to a higher authority). DIGIT offers an auto-escalation mechanism that monitors each complaint’s workflow and advances it if no action is taken within the designated time.

This configuration ensures:

• Timely handling of complaints

• Less manual intervention

• Efficient and automated escalation

---

Configure SLA

1

Open configuration file

Navigate to: data/pg/Workflow/AutoEscalation.json The SLA values are configured here.

Each entry in this file defines:

• Which business service (like PGR)

• Which workflow state (like PENDING)

• What action to perform when SLA is breached

• The SLA duration in days

• And which Kafka topic the escalation event should be pushed to

2

Edit the configuration values (as needed)

Add one entry per state that requires auto-escalation. Fill in fields as shown in the sample configuration.

• Use a realistic SLA value (stateSLA) in days.

  • 1 = 1 day

  • 0.5 = 12 hours

  • 0.01 = ~15 minutes

• Ensure "active" is set to "true" to enable escalation.

---

Sample Configuration

🔹 1. Escalate from PENDINGATLME State

{
  "businessService": "PGR",
  "module": "pgr-services",
  "state": "PENDINGATLME",
  "action": "FORWARD",
  "active": "true",
  "stateSLA": 0.01,
  "businessSLA": null,
  "topic": "pgr-auto-escalation"
}

• If a complaint is in the PENDINGATLME state for more than ~14 minutes (0.01 days), and no action is taken.

• The system will automatically trigger the "FORWARD" action.

• This moves the complaint forward in the workflow (usually to the next responsible officer).

• The message is sent to Kafka topic pgr-auto-escalation for processing.

🔹 2. Escalate from PENDINGFORASSIGNMENT State

{
  "businessService": "PGR",
  "module": "pgr-services",
  "state": "PENDINGFORASSIGNMENT",
  "action": "ASSIGNEDBYAUTOESCALATION",
  "active": "true",
  "stateSLA": 0.01,
  "businessSLA": null,
  "topic": "pgr-auto-escalation"
}

• If a complaint is waiting for an assignee (PENDINGFORASSIGNMENT) for more than ~14 minutes,

• The system auto-assigns it using the "ASSIGNEDBYAUTOESCALATION" action.

• This ensures the complaint does not remain unassigned.

🔹 3. Escalate from REJECTED State

{
  "businessService": "PGR",
  "module": "pgr-services",
  "state": "REJECTED",
  "action": "CANCEL",
  "active": "true",
  "stateSLA": 0.01,
  "businessSLA": null,
  "topic": "pgr-auto-escalation"
}

• If a complaint is in REJECTED state for more than ~14 minutes,

• It will be moved to a CANCELLED status automatically using the "CANCEL" action.

• This helps clean up rejected complaints that are not manually closed.

---

Key Parameters

Parameters	Description
businessService	The module this SLA applies to (PGR in this case)
state	The complaint's current workflow state
action	The next step the system will auto-trigger
stateSLA	Time in days before auto-escalation (e.g., 0.01 = ~14 minutes)
topic	Kafka topic where the escalation event is published
active	Whether this escalation rule is currently active

---

Example Use Case

Imagine a citizen raises a complaint about a streetlight not working:

• It lands in PENDINGATLME state.

• If no one takes action in 14 minutes, the system auto-forwards it to the next level.

• If it's stuck waiting for an assignment, it gets auto-assigned.

• If someone rejects it and doesn’t close it manually, it gets cancelled automatically.