Audio Inference Engineer, Model Efficiency

Advertisements
IT & Technology 🇺🇸 New York 07 Nov 2025
This position has been filled. Explore our other open positions here.

Role overview

Cohere is a security-focused enterprise AI company building foundation models and products for real-world business applications. This Audio Inference Engineer role sits within a growing team focused on optimizing audio model serving efficiency, where you will advance core metrics like latency, throughput, and quality while solving challenges in real-time and streaming audio inference workloads.

Key responsibilities

  • Design and optimize high-performance audio inference systems to improve latency, throughput, and output quality.
  • Identify performance bottlenecks in audio processing and streaming pipelines through deep system analysis.
  • Develop and implement creative solutions for efficient audio processing and real-time streaming inference.
  • Collaborate with training and serving infrastructure teams to ensure smooth integration between model development and production deployment.
  • Work on end-to-end audio pipeline optimization with a focus on real-time and streaming architectures.

Requirements

  • Significant experience building high-performance audio or machine learning inference systems.
  • Strong proficiency in C++ and Python.
  • Hands-on experience developing deep learning models for audio, speech, or language applications.
  • Demonstrated bias for action and results-oriented mindset.
  • Experience with GPU programming and low-level system optimization is highly valued.
  • Background in model parallelization across multiple GPUs, duplex real-time streaming architectures, or machine learning frameworks for audio such as PyTorch or TensorFlow is a strong plus.
  • Familiarity with inference frameworks like vLLM, SGLang, or custom distributed inference systems, and knowledge of sequence modeling for audio and speech are additional advantages.

What makes this role worth considering

You will join a fast-growing team of committed researchers and engineers at a company with global offices across Toronto, Montreal, San Francisco, New York, Paris, Seoul, and London. The Model Efficiency team operates primarily in EST and PST time zones with a remote-friendly culture, offering flexibility in where you work while staying connected to collaborative peers.

Cohere provides comprehensive benefits including a weekly lunch stipend, full health and dental coverage with a separate mental health budget, 401K/RRSP matching, 100% parental leave top-up for up to six months, six weeks of paid vacation, annual enrichment credits for arts, fitness, and learning, and a $500 home office stipend.

How to apply

Use the Apply button to view the official job posting and submit your application on Cohere's careers site.

Advertisements

ImmigrationCafe is a job listing directory, not a recruiter. Applications are submitted on the employer's official careers website.

Location