About the project

B2B automation for airline disruption services. When flights get disrupted, our AI system helps the operations team find and book hotels for stranded passengers.

The QA role covers two phases:

(1) during development — validating Italian voice quality and testing scenarios in the extraction pipeline, and

(2) in staging/production — monitoring AI behavior using observability tools (Langfuse) and running the testing feedback loop.

Objective

Ensure the AI voice assistant speaks natural Italian and follows the correct operations workflow — from initial testing through production monitoring.

Areas of Responsibility

During Development (W3-W6):

Listen to AI-generated Italian voice calls and evaluate: natural pronunciation, correct hotel vocabulary, formal "Lei" register
Review test scenarios for realism
Run test cases, report defects with clear reproduction steps
Work with the operations team (Alessio, Vanna) to confirm domain correctness
Promote test scenarios through quality tiers: draft → reviewed → confirmed

During Staging/Production (W7-W10):

Monitor AI behavior using Langfuse traces — identify patterns, failures, quality drift
Run end-to-end test scenarios against deployed agents
Provide structured feedback for prompt/skill improvements
Participate in UAT alongside the operations team

Skills

Attention to process detail — ability to follow a multi-step workflow and spot errors
Scenario-based testing experience
Comfortable providing structured feedback (not just "sounds wrong" but "the word X should be Y because...")
Comfortable working with observability/monitoring tools (Langfuse — training provided)

Nice to have:

Hotel/hospitality vocabulary in Italian
Basic understanding of AI/automation systems

Languages:

Italian: native
English: advanced (C1) — primary working language

Neurons Lab

Neurons Lab

Neurons Lab