Unisound

Unisound ranks as China’s fourth-largest AI solutions provider by 2024 revenue, with particular strength in healthcare applications where it serves nearly 500 hospitals including Peking Union Medical College Hospital. Founded in 2012 by Dr. Huang Wei and Dr. Liang Jiaen, the company has evolved from a speech recognition specialist into a broad AI platform provider, culminating in its Hong Kong Stock Exchange listing in June 2025 as the “AGI First Stock.”

Company Overview

Unisound emerged from the early wave of Chinese AI startups, positioning itself as one of the first companies in Asia to commercialize deep neural network algorithms for speech recognition. The founding team brought expertise from institutions including Motorola China Research Center and the Chinese Academy of Sciences Institute of Automation.

The company operates through its proprietary UniBrain technology platform, which integrates speech recognition, natural language processing, knowledge graphs, and computer vision capabilities. In 2023, Unisound launched the Shanhai Large Language Model with 60 billion parameters, designed to support dynamic reasoning and cross-modal data processing for human-machine interaction.

Unisound’s computing infrastructure includes an Atlas AI cluster with over 184 PFLOPS of computing power and more than 10PB of storage capacity. The company has also developed its own AI chip series—Hummingbird (蜂鸟) for IoT applications and Snow Leopard (雪豹) for automotive-grade voice AI—with cumulative shipments exceeding 36 million units through 2024.

Core Products

Medical Guide Robot

The Medical Guide Robot serves as an intelligent service terminal deployed in hospital lobbies and outpatient areas. Key functions include department information queries, indoor navigation, symptom-based triage recommendations, medical knowledge Q&A, and basic administrative service guidance. The system supports multiple interaction modes including touch screen, voice commands, and facial recognition.

Healthcare AI Solutions

Beyond hardware, Unisound provides a suite of AI-powered software solutions for medical institutions:

• Voice-Enabled EMR: Real-time speech-to-text transcription for electronic medical records, reducing documentation burden on physicians
• Medical Record Quality Control: AI-assisted review system checking standardization and clinical appropriateness, achieving quality levels approaching human expert review
• Single Disease Quality Management: Specialized quality monitoring for specific disease pathways
• Medical Insurance Payment Management: AI tools for insurance claim review and compliance

Technology & Innovation

Unisound’s technology stack follows a “Chip + Edge + Cloud” architecture strategy. The company develops processors and sensors to enable edge devices, enabling voice-activated interactions while connecting these devices to cloud-based intelligent platforms.

The Shanhai Large Language Model represents the company’s latest advancement, building upon years of iteration on the UniCore model first introduced in 2018 based on BERT architecture. The model achieved a score of 82.2 on the MedBench evaluation benchmark, demonstrating strong performance in medical domain applications.

Key technology areas include:

• Automatic speech recognition with support for multiple Chinese dialects
• Natural language understanding and semantic analysis
• Clinical knowledge graph construction
• Multi-modal interaction (voice, vision, touch)

Market Presence

Unisound’s healthcare business has expanded steadily, with medical revenue growing from CNY 110 million in 2022 to CNY 200 million in 2024, representing an increasing share of total company revenue (18.9% to 21.2%). The company serves 166 healthcare clients as of 2024.

Strategic partnerships began early, with the company establishing collaboration with Peking Union Medical College Hospital in 2016 for voice-activated electronic medical record solutions. Hospital deployments now span across China, with the company ranking fourth in the Chinese medical AI services market with approximately 2.1% market share.

The Smart Life business segment, accounting for approximately 79% of revenue, includes applications in smart home (partnerships with Gree, Midea), automotive (Geely), and transportation (Shenzhen Metro Line 20).

Key Milestones

• 2012: Company founded in Beijing
• 2016: Strategic partnership with Peking Union Medical College Hospital; began building Atlas AI infrastructure
• 2018: Launched UniOne AI chip series; raised $100M in Series C funding
• 2020: Initial IPO filing for Shanghai STAR Market (later withdrawn)
• 2023: Released Shanhai Large Language Model with 60 billion parameters
• 2025: Listed on Hong Kong Stock Exchange (9678.HK) on June 30, raising HK$206 million

Frequently Asked Questions

What medical products does Unisound offer?

Unisound provides a Medical Guide Robot for hospital reception and navigation services, along with AI software solutions including voice-enabled electronic medical records, medical record quality control systems, and healthcare insurance payment management tools. The company serves nearly 500 hospitals across China.

Is Unisound publicly traded?

Yes, Unisound listed on the Hong Kong Stock Exchange on June 30, 2025, trading under stock code 9678.HK. The company raised approximately HK$206 million in its IPO, with cornerstone investors including SenseTime, Zhenyi Asset, and Runjian International.

How does Unisound compare to iFlytek in healthcare?

Both companies compete in China’s medical AI market. According to Frost & Sullivan, Unisound ranks fourth in medical AI services while iFlytek Medical (a subsidiary of iFlytek) reported healthcare revenue of CNY 734 million in 2024, approximately 3.7 times Unisound’s medical segment revenue. Unisound focuses primarily on voice interaction and quality control systems, while iFlytek has a broader medical product portfolio.