Back to use cases
Logistics

Complaints & escalation

The Challenge

A complaint arrives from portal, email or CRM; must be classified, assigned, handled, possibly escalated, closed with evidence.

Common Blockers

  • Unmeasured SLAs: no one knows how many complaints are about to expire until the customer calls back.
  • Manual and informal escalations: the manager is notified via chat only when the situation is already critical.
  • No structured evidence of resolution: in case of dispute, reconstructing the path takes hours.

The FlowSharp Solution

API/event trigger, structured flow, AI task for urgency, manual tasks for customer contact, timeouts to avoid stalling, full history.

Process Stages

  1. 1

    Receipt & ticket opening

    The complaint arrives via API, parsed email or portal form. FlowSharp automatically opens a process instance in state 'Complaint received' and records the source.

  2. 2

    AI urgency classification

    An AI task analyzes the complaint text and classifies it by urgency (low/medium/high/critical) and category. The AI doesn't decide alone: it proposes, a manual task confirms for high-urgency cases.

  3. 3

    Assignment & ownership

    Based on urgency and category, the process automatically assigns to the right team (customer care, operations, legal). An ownership timeout is monitored in real time.

  4. 4

    Handling & customer contact

    The manual customer contact task requires logging the outcome of the call or email. If the customer is unreachable, an automatic retry triggers a new task.

  5. 5

    Closure with evidence & audit

    The complaint closes only when all tasks are complete and resolution evidence is recorded. The immutable history documents every action for potential future disputes.

Expected Outcomes

  • SLAs met: every complaint has a visible timer and automatic escalation before expiry.
  • Zero tickets lost in inboxes: every complaint has an owner, a state and a monitored deadline.
  • Full audit per complaint: who did what, when, with what outcome — available in one click.

Industry note — Logistics

In logistics, a shipment complaint often involves third-party carriers, insurance and liability regulations. FlowSharp keeps the process coordinated even when part of the handling happens outside the company, with manual tasks to collect evidence from third parties.

Frequently Asked Questions

How do I handle a complaint that requires a response within 24 hours?
Configure a 20-hour timeout on the ownership task. If not completed, FlowSharp runs an automatic escalation: assigns to the supervisor and sends an event notification.
Can AI classify complaints in Italian and English?
Yes. LLM models supported by FlowSharp handle multilingual input natively. You can also configure the prompt to classify according to your specific category schema.
Can I connect this flow to my existing ticketing system?
Yes, via webhook or API trigger. FlowSharp can open an instance from an event in your ticketing system (Zendesk, Freshdesk, Jira) and update the original ticket status on completion.

Key Metrics

SLA

Rispettato

Items at risk

Visibili

Features Used

trigger-api
ai-task
manual-task
timeout
events
audit

Ready to govern this process?

Book a demo to see how FlowSharp can implement this exact use case for your team.

Your privacy matters

We use only essential cookies to make this site work. You can also allow optional cookies to help us improve it.