📋 Job Overview

The position focuses on advancing real-time and offline voice content understanding for Yuanbao's 2C and 2B scenarios, enhancing recognition and comprehension quality through techniques like noise-robust speech recognition, rich linguistic information identification, and multi-language support. It involves developing large model solutions for complex audio scenarios including ultra-long audio, extended contexts, multi-speaker interactions, multi-language, dialects, multi-turn dialogues, and continuous speech flows. Additionally, the role emphasizes leveraging multimodal inputs to improve product efficiency and optimizing models for miniaturization, edge computing, and PaaS capabilities to empower audio understanding across Tencent's Yuanbao ecosystem.