Skip to main content

Incident Management Process

Overview​

The AI Cockpit incident management process is designed to ensure agility, traceability, and assertiveness in problem-solving, using a tiered structure (L1, L2, L3) and automatic integration between email, Opsgenie, and Jira.

1. Ticket Opening and Reception​

Tickets must be opened directly through the AI Cockpit Service Desk portal:

👉 AI Cockpit Service Desk

Step-by-step to report an incident​

  1. Access the portal above.
  2. Choose the option to report an incident.
  3. Fill out the form with:
    • Summary: A brief description of the problem.
    • Organization: Inform the affected organization.
    • Detailed description: Include the impacted module (Reasoning, Smart Engineering, Platform, or Metrics), steps to reproduce the problem, evidence (screenshots, logs, videos), context (when it occurred, frequency, impact), and any solution attempts already made.

2. Service Flow​

L1 – NOC (First Response and Triage)​

  • Receives the ticket automatically via Jira.
  • Performs initial validation: understands the problem, checks if it can be handled directly.
  • If it is a simple or known issue:
    • Executes the solution.
    • Documents and closes the ticket.
  • If it requires a more in-depth analysis:
    • Escalates to L2.

L2 – Support (Intermediate Analysis)​

  • Acts as an intermediate level of analysis.
  • Attempts to solve or supplement information on the ticket.
  • If resolved:
    • Documents, attaches evidence, and closes.
  • If unsure how to proceed (especially with questions about processes, rules, or platform functionality):
    • Checks the documentation.
    • Technical questions (infra, logs, unexpected behavior) are normally handled by L2.
  • If the problem persists or is highly technical (bugs, system failures), escalates to L3.

L3 – AI Cockpit Team​

  • Receives tickets escalated by L1/L2.
  • Acts directly on bugs, system failures, and deep technical analysis.
  • Performs the final treatment, implements corrections, and closes the ticket in Jira.

3. Important Points​

  • The standard escalation path must be followed: NOC → L2 → AI Cockpit.
  • All actions must be documented in Jira, with evidence attached.
  • Email with the client should only be used for complementary alignments.
  • Whenever possible, provide as much detail and evidence as possible when opening the ticket to speed up the process at all levels.

4. Process Benefits​

  • Centralization of records and incident history.
  • Diagnosis and correction of failures in the shortest possible time.
  • Optimization of response time and quality of the solution delivered to the user.

📄 Related Articles​