Audio ML engineer

On-site | Warsaw, Poland
Experience: 5+ years in ML/AI
 

Requirements:

• 5+ years of experience in ML/AI engineering with a focus on Generative AI (LLMs, TTS) and voice-based conversational systems.
• Strong background in API/service engineering and streaming implementations.
• Proven experience with hardware/software validation and compatibility checks (GPU, OS, dependencies, audio I/O, codecs).
• Expertise in stress testing, latency consistency, and quality assurance for audio pipelines.
• Experience with model release management: shipping, rollbacks, integrations, and end-to-end impact assessment.

Nice to Have:

• Performance optimization: load balancing, dynamic batching, caching.
• Knowledge of observability and reliability systems (logging/tracing, health checks, alerting, dashboards).

Key Tech Areas: Machine Learning, AI, APIs, Audio Streaming (ASR, TTS, VAD, Diarization)

Responsibilities:

Model Delivery & Integration

• Own end-to-end delivery of audio ML models between Applied Science and Engineering teams.
• Architect and implement streaming audio services (ASR, TTS, VAD, Diarization).
• Validate compatibility across hardware/software setups in development and production.
• Ensure smooth integration with other system components while preserving audio quality and performance.

Testing & Quality Assurance

• Perform stress, soak, and performance testing for audio services.
• Maintain quality, low-latency, and stability under heavy load.

Model Release & Lifecycle Management

• Manage release of new model versions and architectures.
• Provide rollback strategies and monitor end-to-end impact.

Performance & Observability (Optional)

• Optimize performance with batching, caching, and load management.
• Build observability tools: structured logging, tracing, health monitoring, alerting, and dashboards.

Project:

This role is for an Audio ML Engineer responsible for delivering, testing, and scaling audio-based AI systems (speech recognition, synthesis, and conversational AI).