🛠 Tech Stack

💼 About This Role

You'll design and train custom AI models for Channel Talk's products, starting with TTS/STT for voice counseling. You'll build end-to-end ML pipelines and improve speech quality, latency, and efficiency in a small team.

🎯 What You'll Do

• Design and train AI models to solve product problems
• Improve TTS/STT model performance and voice quality
• Build end-to-end ML pipelines from data collection to evaluation
• Collaborate with product engineers to deploy models into service

📋 Requirements

• Experience training TTS or STT models and applying them in service
• Experience training speech models end-to-end with PyTorch
• Deep understanding of Transformer-based speech model architectures
• Experience with GPU training environments (multi-GPU etc.)

✨ Nice to Have

• Experience with multilingual speech models (Korean, Japanese)
• Real-time streaming speech recognition/synthesis experience
• Experience with distributed training (DeepSpeed, FSDP)

📨 Hiring Process

Process: Resume review -> Online tech interview -> 1st interview -> 2nd cultural fit interview -> Reference check -> Final offer.