Company Description

Kardome is at the forefront of voice-enabled AI, pioneering spatial-hearing solutions that redefine hands-free interaction. As part of our next wave of innovation, we seek to develop a wake-word detector that surpasses the accuracy, robustness, and efficiency of every existing solution on the market.

Role Description

This is a full-time on-site role for an AI Researcher located in the Tel Aviv District, Israel. You will lead research and development of state-of-the-art wake-word detection models, pushing the boundaries of what’s possible in low-power, far-field, and noisy environments. Your work will directly impact millions of devices, ensuring instant, reliable activation of voice interfaces even under the most challenging acoustic conditions.

What we are looking for

Someone who will delve deeply into AI models for keyword spotting (and other speech‐processing tasks), oversee every phase of model design and training, and work tirelessly until they outperform all competitors.

Qualifications

• Deep understanding of audio feature extraction (e.g., MFCC, log-Mel spectrograms) and acoustic modeling.
• Expertise in modern neural architectures (CNNs, RNNs, Transformers) and training frameworks (TensorFlow, PyTorch).
• Experience with model compression techniques (pruning, quantization, distillation).
• Proficiency in Python for performance-critical implementations.
• Strong research background
• Excellent problem-solving abilities
• Strong analytical and critical thinking skills
• Master's or Ph.D. in Computer Science, AI, or related field is a plus