Home · Agentic solutions · The complaints agent: answers in hours, not days
Case study · Sales & customers · OperationsCustomers forget the price. They remember the wait.
The complaints agent: answers in hours, not days
The agent reads every complaint from email and forms, gathers evidence from the ERP and TMS, and settles simple cases by your rules. Hard ones reach a human — with a ready recommendation.
Executive summary
2,400 complaints a month in inboxes; each needs data from 3 systems.
The agent classifies, gathers evidence and settles simple cases by rule.
First response in hours; the team handles genuinely contested cases.
The ERP, the TMS, the complaints inbox, Teams; UiPath agent + robots.
A B2B distributor, 2,400 complaints a month
Missing items, transport damage, wrong prices, returns. 65% of cases — 1,560 a month — are simple, decidable from the price list, the delivery note and a photo.
Yet every case waits for a person to open three systems and stitch the order history together. On average 3.2 days to a first answer — while the customer talks to competitors.
The current reality
- HumanThe customer emails; the case lands in a shared inbox
- WaitingWaits to be assigned — a day, sometimes two
- HumanAn agent collects data from the ERP, TMS and email threads
- HumanAsks the warehouse and the carrier what happened
- Error riskOutcomes differ from one employee to the next
- WaitingThe customer hears back after 3 days — and remembers it
The hidden cost of the current process
A complaint is the moment of truth — and yours lasts 3 days.
- 468 hours a month of team time on cases a rule could settle.
- Inconsistent outcomes teach customers that appealing pays — case volume grows.
- Reorders lost to a bad experience never show up in any report.
The cost of doing nothing
Manual work is an operational tax: the automation investment is finite, the manual cost is paid again every month.
The process after automation
- AutomationThe agent reads and classifies the case within a minute
- AutomationGathers evidence: order, delivery, photos, client history
- AutomationSettles simple cases by your rules — credit, return or a reasoned refusal
- HumanContested or high-value cases: a human with a ready recommendation
- AutomationThe customer hears back in hours; the status stays visible
What the automation handles
- Classifying and evidencing every case
- Settling simple cases by rules and thresholds
- Credits and returns executed in the ERP by robots
When a human decides
- Contested, unusual and above-threshold cases
- Strategic clients — the account owner always decides
- Changes to the complaints policy
Before
After
Value model — example assumptions
Business benefits
- An answer in hours rebuilds trust at the worst moment of the relationship
- One rulebook for everyone — fewer appeals and escalations
- The complaints team handles disputes, not copying order numbers
- Complaint data shows what to fix at the source: product, carrier, warehouse
Board-level KPIs affected
What management gains
- A causes dashboard: what drives complaints and what it costs
- Response SLAs measured, not declared
- Early signals of batch, product or carrier problems
Estimate it for your organisation
An illustrative estimate based on your inputs. A model of released capacity — not a savings promise.
Systems in this scenario
Inputs
- Skrzynka reklamacji / formularz
- ERP (zamówienia, faktury)
- TMS / dokumenty dostaw
Mientha agentic layer
- UiPath Agent Builder
- UiPath robots
- Maestro™ · Action Center
Core systems
- ERP (korekty, zwroty)
- CRM
- Teams (decyzje i eskalacje)
Human approval: Teams / Action Center
What we deliver
- A 2–3 month complaints analysis: categories, values, times
- The complaints policy translated into rules and thresholds
- An agent + robots executing credits and returns in the ERP
- A contested-case route with a recommendation for the human
- A pilot on the 2 most frequent categories
- A monthly root-cause report
What we need to start
- A 2–3 month complaints export
- The current complaints policy and value thresholds
- Test access to the ERP and the complaints inbox
- A process owner on the customer-service side
Implementation roadmap
Discovery
We map the process, data and exceptions with process owners.
Design
Target flow, business rules, approval thresholds.
Build
Agents, robots and integrations in your environment.
Validate
Tests on real cases, exception handling.
Go-live
Controlled rollout with human oversight.
Optimise
Monitoring, reporting and continuous improvement.
Typical duration depends on systems and rules — a single process is usually weeks, not quarters.
Risk and controls
Autonomy under control
- Automatic settlements only up to the value threshold you set
- Strategic clients always handled by a human
- Every settlement reasoned and fully traceable
- A kill switch — back to the inbox in a minute
Why now
- Every month is ~€11,700 of work on rule-decidable cases
- B2B customers bring B2C habits: they count response time in hours
- Complaint data is a free map of operational problems — today nobody reads it
Why this matters to:
The moment of truth stops costing you customers.
Root causes surface monthly: batch, carrier, warehouse.
Credits and returns under threshold control — discretion ends.
Questions we usually hear
In contested cases — yes, and that stays. But a customer with a missing pallet does not want a chat. They want the credit today. The agent delivers it.
The agent applies your rules — the same ones your people apply today, just consistently. It refuses as efficiently as it approves — always with reasons.
Photos the agent accepts and checks against delivery records. Physical inspection stays human — but that is a minority of cases, not 100%.
When this may not be the right solution
- A few dozen complaints a month — a tidy inbox and templates will do
- No written complaints policy — rules first, agent second
- Every case needs physical inspection — automation would cover only the workflow
A question for your next board meeting
How many customers placed their last order right after a complaint last year — and who counted?
What does a customer cost you when they reorder somewhere else after a complaint?
Send one real (anonymised) complaint — we will show how the agent would run it.
Send us a sample complaint