About the Role
Develop high-performance inference solutions for audio models with a focus on efficiency, speed, and scalability across diverse deployment environments.
Responsibilities
- Optimize inference pipelines for audio-based machine learning models
- Improve model latency and throughput without sacrificing accuracy
- Collaborate with research teams to implement efficient model architectures
- Profile and benchmark inference performance across hardware platforms
- Develop compression and quantization techniques for audio models
- Support deployment of models in production environments
- Troubleshoot performance bottlenecks in real-time audio processing
- Design low-latency audio preprocessing and feature extraction modules
- Ensure scalability of inference systems under high load
- Integrate models with backend serving infrastructure
- Maintain high code quality and documentation standards
- Work closely with ML researchers to adapt models for efficient inference
- Evaluate trade-offs between model size, speed, and accuracy
- Implement hardware-aware optimizations for CPUs and accelerators
- Contribute to model versioning and deployment workflows
- Analyze memory usage and reduce footprint of audio models
- Support cross-platform compatibility for inference systems
- Develop automated testing for inference correctness and performance
- Stay current with advancements in model compression and inference
- Collaborate on defining best practices for efficient model deployment
Nice to Have
- Master’s or PhD in a relevant technical field
- Experience with speech recognition or audio generation models
- Contributions to open-source machine learning projects
- Prior work optimizing transformer models for inference
- Familiarity with WebAssembly or JavaScript for inference
- Experience with on-device audio model deployment
- Knowledge of acoustic modeling techniques
- Background in distributed inference systems
Compensation
Competitive salary and benefits package
Work Arrangement
Hybrid or remote options available
Team
Part of the core machine learning systems team focused on optimizing inference performance
What We Value
- Technical excellence paired with practical problem-solving
- Ownership of system performance and reliability
- Clear communication across technical and non-technical stakeholders
- Continuous learning and adaptation to new research
Impact
- Your work will directly influence the speed and efficiency of audio AI systems
- Optimizations will enable broader deployment across devices and platforms
Available for qualified candidates
