Position: AI QA Engineer (Voice AI) Company: Aspirity Solutions

**AI QA Engineer (Voice AI) **❗️outside RF and RB, conversational English ❗️

About the role

We value not only code correctness but also the feeling of natural interaction. Our digital employees communicate with people using voice, and any delays, unnatural pauses, or logical errors directly impact the user experience.

We are looking for an** AI QA Engineer** who will systematically test voice agents and verify their ability to listen, process information, and speak in real-time.

Responsibilities

• Voice UX Testing

• Latency measurement and analysis (Latency P95/P99).

• Barge-In functionality check (correct reaction to interruptions).

• Testing performance in noisy environments.

• Logic and scenario verification

• Reading and analyzing NSL (Jinja2) scripts.

• Identifying logical loops and dead ends.

• Verifying correct state transitions (e.g., booking created → confirmation sent).

• Hallucination control

• Identifying deviations from instructions and incorrect fact generation.

• Formation and maintenance of Golden Datasets for regression testing.

• Integration testing

• Verifying correct calls to external services (CRM, API, etc.).

• Analyzing agent behavior during API errors.

• Verifying correct exception handling.

Requirements

• Technical literacy

• Ability to read code (Jinja2, JSON, Python).

• Confident work with JSON payloads.

• Knowledge of Voice AI

• Understanding of STT (Speech-to-Text) principles.

• Understanding of TTS (Text-to-Speech).

• Understanding of VAD (Voice Activity Detection).

• Awareness of the importance of correct pause and silence handling.

• QA mindset

Languages

Russian — native / fluent.

English — B2 or higher (fluent spoken communication).

✅ Will be an advantage

• Experience with LLM (prompt engineering).

• Experience testing voice assistants (Alexa, Google Assistant, etc.).

• Understanding of AI system quality metrics.

We offer: Remote work format. Opportunity to develop in the QA field for systems with LLM and AI agents. Payment based on qualification: from 20 USD/hour.