Job Description

Speech AI Research & Model Development

• Explore and prototype new approaches for speech and audio AI problems.
• Develop, fine-tune, and adapt speech models for real-world use cases.
• Evaluate modeling approaches and translate promising research into product-integrated solutions.

Speech AI Systems & Audio Pipelines

• Develop and optimize voice technologies, including ASR, TTS, audio processing, speech enhancement, speech understanding, and voice interaction pipelines.
• Design and optimize production-ready audio AI pipelines, including audio preprocessing, feature extraction, noise suppression, real-time audio streaming, model serving, inference pipelines, and cloud/mobile model integration.

AI Productization & Product Experience

• Collaborate with Product, Mobile, and Backend Engineers to build voice recording, transcription, AI-powered voice assistants, voice-driven workflows, and conversational experiences.
• Move AI capabilities from prototype to reliable production systems.
• Improve speech recognition accuracy, speech synthesis quality, inference latency, resource consumption, scalability, reliability, and real-world model performance.

Job Requirement

Education & Experience

• 3+ years of experience building AI/ML systems or AI-powered products; hands-on experience deploying AI models into production.
• Preferred: born from 1997–2003, graduated from top technology universities such as HUST or VNU, good graduation ranking or top graduate profile, and experience building large-scale voice products.

Skills & Competencies

• Strong Python programming skills.
• Experience with deep learning frameworks such as PyTorch or TensorFlow.
• Solid understanding of speech processing and audio machine learning.
• Experience with ASR, TTS, Audio AI, or Voice AI applications.
• Ability to design and implement end-to-end AI workflows.
• Strong research capability: reading, mapping, and customizing research into practical solutions.
• Strong problem-solving skills with a product-oriented mindset.
• Experience researching, reproducing, and evaluating state-of-the-art Speech AI methods.
• Ability to design experiments, benchmarks, error analyses, training/evaluation datasets, synthetic data, augmentation, and quality validation.

Language / Other Requirements

• English communication is not heavily required.
• English must be sufficient to read and research technical documents.
• Nice to have: Whisper, wav2vec2, NVIDIA NeMo, XTTS, VITS, SpeechBrain, Kaldi, or equivalent technologies.
• Strong advantage: AI deployment on mobile devices, real-time audio streaming, speaker recognition, speaker diarization, voice cloning, audio signal processing, LLM integration, or conversational AI systems.

Personality & Attitude

• Enjoys solving challenging AI problems.
• Product-oriented and focused on real business impact.
• Research-minded, experimental, and able to continuously improve models based on production feedback.