QA Case Study · AI Testing · Verified Demo

Testing an AI Customer Support Center: when the UI says “safe” but the workflow disagrees

This project began as a polished AI Studio prototype. It became a quality-engineering laboratory after testing exposed identity leakage, contradictory routing, misleading escalation promises and a deployment architecture that did not yet match its claims.

Educational portfolio simulation. Do not enter real customer, payment, health or account data. No real support agent is contacted.

6/6scenario contracts passed
3regression rounds
0client-side API secrets
Quality engineering pipeline from requirements to verified deployment
The release gate uses evidence from requirements, audit, remediation, regression and deployment checks.

1. What was built

The prototype models an AI-assisted support centre with six predefined customer scenarios and a free-form testing playground. It classifies an inquiry, assigns category and priority, extracts entities, retrieves a verified FAQ, proposes a response and decides between automated guidance and a simulated human queue.

The browser client is built with React and Vite. The public deployment uses Netlify Functions for the analysis endpoint. Gemini can assist the classification, while a deterministic engine provides a transparent fallback. Both paths must return the same normalized response schema. Demo tickets and FAQ edits remain in the visitor’s browser and are automatically cleared after 24 hours.

2. The real testing problem: defining an oracle

Fluent text is not a pass criterion. Each guided scenario therefore received an explicit contract: expected category, priority, route, FAQ identifier, required entities and prohibited entities. This turns an open-ended AI response into something testable without pretending that every free-form customer message has one perfect answer.

Risk-based coverage focused on state isolation, conflicting truth sources, low-confidence decisions, safety wording, entity semantics, input boundaries, dashboard metrics and deployment claims.

3. What the first audit found

FindingWhy it matteredInitial status
Sarah was addressed as “Marcus”; Elena as “Robert”Cross-request state contamination is a trust and privacy defect.Major
Scenario card said Automated FAQ while runtime used Human QueueTwo sources of truth produced contradictory behavior.Major
Low FAQ match displayed both automated and human decisionsA user could not tell which route was actually selected.Major
Simulation promised a manager would callThe demo implied a real human action that could not occur.Major
Express API was described as Netlify-readyA static deploy would not automatically host the backend.Major
Escalation KPI and status filter used different semanticsSummary numbers could not be reconciled with detailed records.Minor

Negative tests also covered empty and whitespace input, HTML/XSS-like strings, low FAQ relevance and long transcripts. The tested XSS-like input rendered as text, but that single check was not presented as comprehensive security proof.

4. How the system was corrected

  • Customer identity was bound to the active request and previous names were forbidden in the prompt and normalized response.
  • One normalized decision object became the source for UI badges, routing, FAQ explanation and ticket creation.
  • All six guided scenarios received automated contracts instead of isolated hard-coded fixes.
  • Entity rules separate order and invoice IDs from product SKUs and omit absent values instead of rendering raw null.
  • Input is limited to 4,000 characters in the UI and function, with an explicit 400 response.
  • Escalation language is hypothetical and visibly labelled as a simulation.
  • The Express-only path was complemented by Netlify Functions, redirects, request validation, rate limiting, deterministic fallback and server-only Gemini credentials.
  • Browser data now has automatic 24-hour expiry, plus manual Reset Demo.
Six verified support scenario contracts
Final contract regression: all six guided scenarios passed category, priority, route, FAQ and entity checks.

5. Regression was a separate product phase

The first regression confirmed several fixes but reopened Delivery routing and found missing order extraction, raw null values, a 4,001-character boundary failure, an inaccurate localStorage statement and a badge mismatch. The second focused round closed most of those issues but detected a new regression: the damaged-item refund had drifted from Billing and Payments to Delivery, and the SKU appeared as an order number.

The third round verified the corrected normalized contracts in different scenarios. All six passed. This chronology is intentionally preserved: the audit report is not overwritten by the later PASS result, because the defect history is part of the evidence.

ScenarioExpected contractResult
Delayed DeliveryDelivery · Medium · Automated FAQ · US-98214PASS
Damaged Item RefundBilling and Payments · High · Human Queue · SKU-4401 · $89.50PASS
Duplicate BillingBilling and Payments · High · Human Queue · INV-7731 · $49.00PASS
Mobile App 503Technical Support · High · Automated troubleshootingPASS
Angry ComplaintComplaints · Critical · Human Queue · Manager EscalationPASS
Support HoursGeneral Questions · Low · Automated FAQ · no invented entitiesPASS

6. Deployment is part of QA

A successful AI Studio preview does not prove a Netlify build. The release gate therefore checks the actual Functions routes, Gemini success and deterministic fallback, HTTP 400 and 429 behavior, CSP and security headers, absence of secrets in client assets, browser-data expiry, cold-start behavior and desktop/mobile rendering.

The Gemini key is read only inside the serverless function. The frontend never receives it. CORS is not opened with a wildcard because the application calls its API on the same origin.

7. What this demonstration does not claim

  • It is not a real customer-service operation and does not contact agents or process refunds.
  • The in-memory function limiter is suitable only for a low-traffic educational demo, not a distributed production service.
  • Microphone start/stop was tested; full speech-to-text accuracy was not certified with a controlled audio corpus.
  • Guided scenario contracts stabilize demonstrations. Free-form playground inputs remain intentionally open to evaluation.

8. Reusable lessons for AI testing

  1. Test the same request after every other scenario to expose stale context.
  2. Assert contradictions across fields, not only whether each field looks plausible.
  3. Separate model judgment from deterministic business and safety rules.
  4. Treat safety copy, persistence statements and deployment claims as testable behavior.
  5. Keep the original failures and later regression evidence separate.

Try the controlled demonstration

Use fictional data and observe the normalized decision, matched FAQ, entity extraction, route and execution engine. The interface clearly identifies Gemini-assisted and deterministic modes.

Open AI Customer Support Center ↗

Related work

Comments

Leave a comment

Share your thoughts about the case study. No account is required.

Comments

Loading comments…

Comments are published immediately and can be moderated later.