E
Senior Data Scientist – Audio and Speech AI
במקום העבודהמשרה מלאהData ScienceDefense and Space
Develop advanced audio, speech, and multimodal AI systems for real-time understanding, generation, and analysis in noisy operational environments. The role covers research, model development, evaluation, and on-premise or edge deployment.
תחומי אחריות
- Fine-tune and evaluate speech-to-text models for noisy, low-latency, mission-critical environments.
- Develop speaker identification, diarization, sentiment, and emotion analysis systems.
- Design and optimize multimodal pipelines combining audio, text, and visual inputs.
- Develop generative audio capabilities including noise reduction, voice conversion, speech enhancement, and conversation insights.
- Collaborate with machine learning engineers and research teams to deploy, scale, and optimize models on-premise and on edge hardware.
- Adapt models for real-time speech understanding, decision support, and behavioral insights.
דרישות
- At least 5 years of hands-on experience developing and deploying speech or audio AI models.
- At least 3 years of experience with STT/ASR, TTS, speaker recognition, or sentiment analysis.
- Strong background in machine learning, deep learning, and audio signal processing.
- Experience with noisy real-time audio, latency optimization, and edge-device constraints.
- Familiarity with Conformer, Whisper, RNN-Transducer, FastSpeech, Tacotron, speaker embedding networks, and self-supervised speech representations.
- Understanding of semantic embeddings, multimodal search, and retrieval-augmented generation architectures.
- Experience with Agile workflows, MLOps, and DevOps principles.
- Strong data-driven approach and ability to research novel audio AI methods.
יתרון
- Publication record, Kaggle participation, challenge participation, or equivalent achievements.
הטבות
- Collaborate with researchers and engineers on next-generation speech and audio intelligence.
- Work on multimodal systems integrating vision, audio, and language.
- Contribute to real-world speech understanding, generation, and sentiment analytics.
- Shape solutions from research and concept through deployment.