Customer support

SLA Breach Forecaster

Names the tickets that will breach, while there is still time

Version 2Published Sep 2, 2026
Customer supportKnowledgeExtractclaude-sonnet-5Read-only · auto8 hard rules5 handover rulesResult
8hard rules
Things it must never do, enforced in every run
5handover rules
Conditions where it stops and gives the work to a person
10output fields
A structured result, not a paragraph you have to re-read
≤6steps per run
Bounded work: at most 5 tool calls, enforced by the runtime

Read-only by design. This agent has no path to sending, spending, publishing or committing anything. Where that limit is the product, removing it would remove the reason to trust the output.

Live wiring for version 2. Tool nodes are the capabilities this agent is registered to call — nothing else can run.

Overview

Every service desk reports breaches after they happen. The standard dashboard shows a percentage that is already history — a scoreboard, and a scoreboard cannot be acted on. By the time a ticket appears in the breach count, the credit is owed and the only remaining option is an apology.

What a service manager needs is the list of tickets that will breach if nothing changes, ordered by time remaining. Almost no tooling answers that, because it means predicting how long a ticket will take rather than measuring how long it took.

The prediction does not need to be clever. Most breaches are visible hours ahead from a handful of facts: the ticket is waiting on a third party who has not replied, it has been reassigned twice, it is the only one of its kind assigned to someone who is out tomorrow, or it sat untouched over a weekend the clock ran through.

The second thing nobody checks is the clock itself. Pause rules, business calendars and time-zone handling are where SLA measurement quietly goes wrong, and a misconfigured pause rule means your reported figure and your contractual one differ — discovered only when a customer disputes a credit.

This agent reports both: what is about to breach with the time left and the reason, and where the measurement itself does not match the contract. It reassigns nothing and answers nothing.

Benefits

Breaches prevented rather than reported

Which is the only version that helps a customer.

Credits kept

A prevented breach is money that never leaves.

Measurement you can defend

Because the clock rules were checked against the contract.

What it does

  • Forecasts, rather than counting

    A breach percentage is history. A list of what breaches next is something a person can act on.

  • Gives the reason, not just the risk

    Waiting on a third party, reassigned twice, owner away tomorrow. Each needs a different intervention.

  • Checks the clock against the contract

    Pause rules, calendars and time zones. A misconfigured pause means your figure and the customer's disagree.

  • Prices the exposure

    The service credits at stake, so priority reflects consequence rather than age.

  • Finds the tickets nobody owns

    Reassigned repeatedly, or sitting with someone who is out. The classic breach, entirely preventable.

  • Touches no ticket

    It never reassigns, reprioritises or replies. Reordering someone's queue is their manager's call.

What this changes

What this changes about your week

It takes over
Working out which open tickets are about to breach, and why.
Done today by
A service manager scanning a queue, usually after the fact.
How often
Daily. The window to prevent a breach is hours.
Typical manual effort
Every ticket needs elapsed time, waiting state, owner availability and the clock rules held together.

Why Customer support in particular. Every service desk reports a breach percentage that is already history. And nobody checks the clock itself — a pause rule the contract does not support means your figure and the customer's disagree.

  • Runs unattended

    Started by you or by an event, and it finishes on its own. Nothing waits for someone to be at a desk.

  • The same standard every time

    The two-hundredth item is held to the bar the first one was. Consistency is the part people cannot sustain.

  • It cannot act on its own

    SLA Breach Forecaster has no path to sending, spending or committing. That limit is why its output is safe to act on.

How it runs

This agent runs server-side through the PROMIVO runtime. Each run is logged step by step and every tool call is permission-checked before it executes.

  1. 1InputValidated against the agent input schema
  2. 2Modelclaude-sonnet-5 with automatic fallback
  3. 3Toolssearch_knowledge, extract_structured_data
  4. 4PermissionsRuns READ actions automatically; anything higher waits for approval
  5. 5OutputStructured result matching the output schema
  6. 6LoggingTokens, duration and every step recorded

Demo

Demo dataIllustrative sample output, abridged.

Input

{
  "asAt": "2026-09-02T09:00:00Z",
  "tickets": [],
  "warningWindowHours": 4
}

Output

{
  "escalate": true,
  "disclaimer": "A forecast from your own records and contract terms. No ticket has been changed or answered, no credit issued or promised, nobody contacted, and no statement is made that a contractual breach has occurred.",
  "willBreach": [
    {
      "basis": "Elapsed against target, with the pause rule corrected to the contract.",
      "reason": "Waiting on a third-party supplier who has not replied for 3 days. The clock is not paused for third-party waits under this contract, though the tool shows it as paused.",
      "customer": "Harbour Logistics",
      "reference": "INC-8841",
      "creditExposure": 4200,
      "hoursRemaining": 2.5
    }
  ],
  "clockFindings": [
    {
      "issue": "Pause on third-party wait",
      "contract": "Clause 7.4 pauses only for waits on the customer, not on your suppliers.",
      "configured": "Clock pauses whenever state is 'Waiting — third party'.",
      "ticketsAffected": 19
    }
  ],
  "couldNotAssess": [
    "Three tickets have no target recorded."
  ],
  "alreadyBreached": [
    "INC-8790 breached 6 days ago. Clause 11.2 makes a credit automatic and no credit appears on the account."
  ],
  "escalationReason": "A pause rule is applied that the contract does not support, affecting 19 tickets, and one ticket breaches within 2.5 hours.",
  "unclaimedCredits": [
    "One breach from last month appears to owe a credit that was never applied."
  ],
  "contractTermsFound": true,
  "totalCreditExposure": 18400
}

Requirements

Integrations

No integrations required.

Before you start

  • Open tickets with target, elapsed time, state and owner
  • Your SLA terms, including pause rules and business calendar
  • Service credit terms, where they exist
  • Requires the All-Access plan or higher

Use cases

Daily service review

What breaches today and tomorrow, and why.

Before a customer review

Exposure and cause, ahead of the meeting.

After a contract change

Whether the tooling actually implements the new terms.

Holiday and cover planning

Which commitments depend on who is in.

Pricing

All-Access

Billed monthly through your PROMIVO subscription. Cancel at any time.

Runs consume your plan allowance for agent executions and tokens. See plan limits

FAQ

Will it reassign or escalate tickets?

No. Reordering someone's queue is a management decision with context this cannot see. It produces the list and the reason.

How does it predict a breach?

From facts already in the record — elapsed time against target, waiting state, reassignment count, owner availability, and the clock rules. It states the basis for each and never presents a guess as a measurement.

What if our SLA configuration is wrong?

That is one of the findings, and often the most valuable. Where the configured clock does not match the contract terms supplied, it reports the difference rather than trusting the tool.

Reviews

Reviews open once customers have run this agent.

Need this agent to work differently?

Tell us what to change and our team will scope a customised version for your business.

Customize this agent