Job Title: NLP/Audio ML Engineer – ASR, TTS & TransformersLocation: Remote Employment Type: Full-TimeAbout Us:Join a cutting-edge AI company pushing the boundaries of speech and language processing. We’re developing scalable, production-ready solutions in Automatic Speech Recognition (ASR), Text-to-Speech (TTS), and advanced Transformer-based NLP. If you’re passionate about deep learning, voice tech, and building state-of-the-art systems, we want to hear from you.Key Responsibilities: • Design and develop sequence-to-sequence models with attention and Transformer architectures • Build and fine-tune encoder–decoder frameworks using TensorFlow Keras or PyTorch • Work with state-of-the-art ASR/TTS toolkits (e.g. ESPnet, Fairseq S2T, Tacotron, WaveNet) • Implement audio feature extraction pipelines (e.g. MFCCs, spectrograms, mel-filterbanks) • Collect, clean, and align large-scale parallel corpora (text and/or speech) • Evaluate models using metrics like BLEU, TER, MOS, and word-error-rate • Optimize and deploy models via Docker/Kubernetes, serving over REST or gRPC • Apply model compression techniques (quantization, pruning) for efficient inference • Customize Transformer models (BART, T5) with advanced techniques like novel attention heads and positional encodingsWhat You Bring: • Strong knowledge of deep learning for NLP and speech • Hands-on experience with large-scale model training and deployment • Familiarity with cloud-based ML environments and scalable infrastructure • A collaborative, solutions-oriented mindsetBonus: • Contributions to open-source projects or publications in ASR/NLP/TTS • Experience in real-time inference and streaming architectures
Job Title
NLP/Audio ML Engineer – ASR, TTS & Transformers