Tickets

Tickets are the heart of BareNOC — every request, alert, or task becomes a ticket tracked from open to resolution by the Queue Manager, AI Technician, and Human Tech roles.

Lifecycle

flowchart LR
    A["Open"] -->|"AI Tech picks up"| B["In Progress"]
    B -->|"needs approval / blocked"| C["Escalated"]
    C -->|"approved"| B
    B -->|"action done — ask customer"| D["Customer Action"]
    D -->|"customer confirms / replies"| B
    B -->|"ok-to-close"| E["Closed"]
    C -->|"human closes"| E
    D -->|"human closes"| E
StatusMeaningWho moves it
OpenNew, not yet picked upcreated by user/alert
In ProgressAI Technician workingAI
EscalatedWaiting on a human tech (approval, guidance, or failure)AI
Customer ActionWaiting on the customer (feedback, info, or closeout)AI
ClosedDone — by a human, or by the AI on customer confirmationhuman / AI

Closure rule: tickets are closed by a human (tech or customer) or by the AI when the customer confirms the outcome ("ok to close").

Priorities

PriorityMeaningHandling
P1Outage — critical service downHighest; escalated immediately
P2Major issue, workaround possibleHigh; fast-tracked
P3Routine requestStandard queue
P4Minor / cosmeticLowest priority

Opening a ticket

  • Web app: Tickets → New Ticket — title, description, priority, device.
  • Chat client: just tell the Queue Manager the problem — it opens a ticket (or appends to the one you're discussing) and confirms with the ticket ID:
    • "the internet is down"
    • "Please tag production and storage vlans on the interface to 192.0.2.64"
    • or explicitly: /new !P2 internet is down | details
flowchart LR
    MSG[Message] -->|"queue/ticket question"| ANS[Queue Manager answers]
    MSG -->|"anything else"| TK[New or appended ticket]
    TK --> AI[AI Technician]
    AI -->|"answers / acts"| FEED[Customer Action — verify]
    FEED -->|"confirmed"| CLOSE[Closed]

Tracking

  • Chat: /status TKT-…, "what's the status of TKT-…", or "are there any notes yet?" (about the ticket in context).
  • Web: the ticket view shows the full conversation (work notes), the action the AI chose, confidence, model, and resolution.
  • Queue views: "what tickets are open", "what's escalated", "what needs customer action", "closed tickets".

Approvals & escalations

How much needs a human at all is set per deployment in the Autonomy Policy — autonomous (home) runs approved actions itself; strict (Tier II) puts every write + P1/P2 in the approval queue.

  • Actions that need a human (reboots, patches, port changes, anything unusual) move the ticket to Escalated — review and Approve, Retry, or Reject/Close from the ticket view.
  • If the AI needs your input, the ticket moves to Customer Action with its question in the thread — reply right there and the AI continues.

AI Technician notes

Every step is recorded in the ticket's work notes with timestamps and actor: processing, llm_response, auto_execute, agent_completed, ai_tech_feedback, escalated, customer_input, completed (close).