Digital Business Services (DBS) Agent

DBS Agent is an AI-powered assistant that automates IT service, incident, and operations management through natural conversation and intelligent workflows.

It delivers faster resolutions with lesser manual intervention. It also improves SLA compliance.

Use Case 1 : Initiating a Request with the IT Support Agent
  • Scenario : User opens the IT Support Agent chat window and types the request -, “Hi, I want to install Notepad++”.
  • Action : The agent acknowledges the request and ask for additional details, if required (such as special configurations or plugins).
    Figure 1. DBS Agent - Agent's action on an Incident

The agent will confirm your request and ask whether to proceed with a standard installation or with specific configurations.

The user types their response (e.g., “Proceed with standard one”).

The agent then automatically creates a Service Request in ServiceNow and shares the Request ID.
Figure 2. Agent's action on an Incident (Cont.)

The designated approver (e.g., manager or IT admin) receives a Service Request Approval Notification.

The notification includes:
  • Ticket ID
  • Request Description
  • Action buttons: Approve or Later
The approver can review the request details and approve it directly within the chat interface.
Figure 3. Agent's action on an Incident (Cont.)

Once approved, the agent updates the user:

“Your Service Request REQ0017646 for Notepad++ has been approved and completed.”

The user can ask for the latest status at any time. The agent provides real-time updates and confirmation upon installation completion by the IT team.
Figure 4. Agent's action on an Incident (Cont.)
Use Case 2.a : Automated Incident Trend Analysis

  • Scenario : The agent begins analyzing all incidents from the last 30 days.

  • Action : The agent performs the following automated steps:

  1. Data Retrieval – Collects incident data (ID, summary, notes).

  2. Data Filtering – Removes test/system incidents.

  3. Pattern Recognition – Clusters issues by trend and automation potential.
    Figure 5. Agent's action on an Incident (Cont.)

The agent displays 20 cluster IDs (CL-001 to CL-020).

The user selects a cluster to analyze (e.g., CL-003). This selection triggers a deep-dive report into recurring issues and related incidents.
Figure 6. Agent's action on an Incident (Cont.)

The agent generates a detailed analysis for the chosen cluster.

The report provides:
  • Problem: (Example: Email delivery failures - SMTP timeout).
  • Automation Candidacy Level (e.g., Partial).
  • Problem Management Recommendation.
  • Related Incident IDs, statuses, and resolutions.
The analysis identifies systemic problems and common failure patterns.
Figure 7. Agent's action on an Incident (Cont.)

The agent maps incidents to KB articles or generates SOPs (Standard Operating Procedures) using GenAI.

Each incident row shows:
  • Status (Open, Pending, Resolved).
  • Linked KB articles.
  • Option to Generate Script (Ansible/PS/VB) or Generate SOP.
The user can view, regenerate, or update mappings as needed.
Figure 8. Agent's action on an Incident (Cont.)

The agent creates a Standard Operating Procedure (SOP) for issue resolution.

Example SOP: Resolving Email Delivery Failures (SMTP Timeout)
  • Identify impacted users
  • Check SMTP service
  • Clear mail queue
  • Verify network connectivity
  • Confirm delivery success
  • Communicate resolution
    Figure 9. Agent's action on an Incident (Cont.)

The agent allows creation of automation scripts using Ansible, PowerShell, or VBScript.

The user selects Generate Script or proceeds to Generate Business Case (Step 6).

This step transitions the workflow from problem analysis (SOPs, KB mapping) to the implementation of automation assets.
Figure 10. Agent's action on an Incident (Cont.)
Use Case 2.b : Generate Automation Business Case

  • Action : The agent compiles performance and effort metrics into a Business Case Report.

The report displays:
  • Total incident volume
  • Average resolution time
  • Manual effort saved
  • Automation success rate
  • Annualized cost/time savings (e.g., $24,240/year)
    Figure 11. Agent's action on an Incident (Cont.)

Once reports, SOPs, and scripts are reviewed, the user selects “End Analysis”.

This action finalizes the session and logs all outputs for audit and reference.
Figure 12. Agent's action on an Incident (Cont.)
Use Case 3 : Identify Open Incidents from Local Database

  • Scenario : The agent queries the local database to find incidents not marked as Resolved or Closed.

  • Action : The agent displays the list of open incidents with details like Incident ID, Status, Start/End Dates.

User is prompted: “Would you like to proceed to Step 2: Incident Data Synchronization?”
Figure 13. Agent's action on an Incident (Cont.)

The agent connects to the ITSM tool to validate each open incident’s status and SLA fields. It checks if any incident has changed to Resolved or Closed and enriches missing data (start/end dates) automatically if the incident is still active.

The output summary shows the validation result (e.g., "Both incidents remain Assigned (Active). No updates or enrichment required.").

User is prompted: “Proceed to Step 3: SLA Risk Assessment & Escalation?”
Figure 14. Agent's action on an Incident (Cont.)

The agent retrieves SLA reports from ITSM for each incident.

It calculates the % Time Spent versus the planned SLA window.

Incidents exceeding a threshold (e.g., >15% time spent) are flagged as breached and escalated automatically.
Figure 15. Agent's action on an Incident (Cont.)

The agent polls the event management system to list critical events requiring attention.

The agent displays event details like severity, entity, and timestamps.
Figure 16. Agent's action on an Incident (Cont.)
The agent analyzes actionable events, performing a deep-level triage to determine the following:
  • Business context
  • Potential root cause hypothesis
  • Recommended runbook or automation
  • Assigned resolution group
    Figure 17. Agent's action on an Incident (Cont.)

Agent logs newly triaged events into the local incident database.

Updates include incident IDs, statuses, and timestamps.

The user can now:
  • End the analysis
  • Restart the workflow
  • Or proceed with specific event triage or reporting actions
    Figure 18. Agent's action on an Incident (Cont.)