AboutLanghui.
Established in Changsha in 2011 for self-study LangData PlatformThe government is not going to be able to get the money.
Quality training data infrastructure support for front-line artificial intelligence development
Professional
AI Data Infrastructure Service
Langhui Founded in Changsha in January 2011It's a self-study. LangData Process PlatformProvides core-driven access from data sources to AI training data set outputsWhole house data serviceHigh-tech enterprises.
We are committed to providing high-quality "training data" support for the development of enterprise artificial intelligence, government artificial intelligence, and creative artificial intelligence.Define the quality of service by indicating precision，Professional capacity-building with specialist depths• Accelerate technological innovation and deployment.
Key building blocks driving the AI data infrastructure
From a mission, with a vision at the helm, with values anchoring every step forward
Our Mission
Build a high-quality AI data infrastructure and lower the threshold for industrial intelligence。 We believe that every well-specified data is laying the foundations for the reliable landing of artificial intelligence.。
Our Vision
As a company that runs healthy, sustainable development.
Our Values
Data quality, lifeline；Expert wisdom driven precision；Open collaboration for value creation。 He insisted on "accurately defining the quality of service and building professional capacity in depth."。
Nine major areas 60+ dataset 200 million+ entries
Coverage of health, medicine, health care, international languages, scientific education, law, world model, industrial economy, multi-areas of life
Healthcare
34 data sets: 50 million electronic cases, 5 million chest X-rays, 5 million head CT, 5 million endoscope videos, pathology WSI, electrocardiograms, skin lenses, eye colours, etc.
Pharmaceuticals
6 data sets: structured 5 million + tablets, 2 million + tablets of Chinese medicine, 1 million + bars of medical equipment, covering ATC 5 full code
Medical Insurance
4 data sets: 10 million + medically insured settlements, DIP/DRG cluster code, health-care catalogues, rights-based data set, 14 + core dimensions
International language material
1 data set: 10 languages 5 million + multi-lingual medical texts, ICD-11/SNOMED-CT terminology, quality of S/A translation
Research & education
5 data sets: 10 million university science topics, 2 million high-quality subject banks, examination data, dissertation data, and scientific and technical AI data
Law
2 data sets: 2 million legal files (Law of the Sea Department), 550,000 legal pleadings
World model/spatial data
6 data sets: 7.25 million photo modelling, 16 million drone modelling, laser cloud 1TB, 3D model split, etc.
Industrial economy
5 data sets: CAD drawings, contract data, financial data, solutions, supply chain data
Public Livelihood
5 data sets: house tickets/placement, expropriation, legal cases, legal contracts, government public service data
LangData platform Six-step full-chain data service
From raw data collection to AI training data delivery, full process closed loop control, end-to-end quality assurance
LangData Data Processing Platform
Data acquisition, cleansing, formatting standardization, smart labelling, quality control, delivery of a training-based AI data infrastructure platform, Provide end-to-end data solutions for health, government, education, etc.。
Data acquisition and de-identification
Covers the collection of multi-source data such as electronic medical records, medical images, clinical studies, etc.; de-identification processing, removal of direct identifiers, virtualization processing of sensitive information such as rare diseases, etc.
Data preprocessing and standardization
Data cleaning (deduplication, error correction, rejection of blurred images), format conversion (uniform DICOM resolution, ICD-11 standard encoding) to ensure data consistency and availability.
Rule-setting for labelling
Medical experts formulate labeling guidelines to clarify entity identification standards (including full names, aliases, and ICD codes), and labeling teams master the use of bio labeling methods and professional tools through case drills.
Multistate label execution
Text NER annotation (symptom, drug entity + relationship extraction), image polygon annotation (tumor boundary, lesion classification), bone key point positioning, covering full modal data.
Level 3 quality inspection and clearance
The marker self-test (accuracy rate ≥ 95%) sampling test (recall rate ≥ 90%) of the→ quality inspection group is finalized by→ medical experts. Disputed cases are finalized by multidisciplinary consultation.
Data delivery and modelling training
Standardized dataset output to support direct access training for a variety of AI frameworks. Provide data version management, incremental updates and continuous optimization services to ensure model iteration efficiency.
Core data service capacity
Focus on the AI data service for 15 years, four core service capabilities to protect your AI project.
Medical data label (total pattern)
Covering full-mode medical data such as CT/MRI/X-ray/ultrasound/DICOM images, electronic medical record texts, ECG signals, endoscopic videos, etc., level 3 quality inspection + final review by medical experts, aligned with China's Personal Information Protection Law, Data Security Law, and project-specific privacy agreements (no HIPAA certification claim).
For more informationSTEM Production of Subjects (Medical Science and Chemical Science)
10 million high-quality science subjects above the university, covering the whole disciplines of mathematics, physics, chemistry and biology, supporting text + picture multimodal question types, suitable for education model training and evaluation.
For more informationLong text factual training data
Covering medical literature, legal documents, academic papers and other long-text scenarios, providing high-quality training data such as entity relationship annotation, logical chain annotation, and fact consistency verification, empowering large models to understand and generate long-text.
MVD neurosurgery label
1 million + annotated surgical video frames, pixel-level semantic segmentation of 24 types of anatomy and surgical instruments, supporting high-precision neurosurgical annotation scenarios such as microvascular decompression (MVD), providing professional training data for surgical navigation AI.
For more informationNational network of molecular companies
1 Group headquarters · 3 holding subsidiaries · 6 branches, covering 10 provinces and cities across the country
Group HQ
HQChangsha Langhui Information Technology Co., Ltd.
Date of establishment: 1 November 2011
Zhang Shagau's new development area, No. 27, Shibuya Park F1, F2 House 2006
• National Operational Headquarters
Subsidiaries
3Shaanxi Langhui Information Technology Ltd.
Founded in December 2022
In collaboration with Innovation Port, New East City Red Light Boulevard, Siam Xinxi Province
Hunan's information technology
Founded in 08/2020
Zhang Shagau's new development area, 27th Shibuya Park F1
Hunan Tobi Technology Ltd.
Founded in February 2020
80 Yingdu Road, Huannan's new district.
Branches
6 Under the banner of Chang Shalang HuiChengdu branch
Founded in 2005
399 East of Chengdu City Capital.
Chongqing branch
Founded in June 2018
Midway to Wongshan Boulevard in Chuking City
Jiangxi Division (Ping Town)
Founded in 2019, 03.
Technological Development Zone, Pingzhou City
Jilin branch
Established in November 2018
Zhangchun City Net Moon Development Area
Guangxi branch
Established in November 2018
Yulin City, 2nd Rim North Road, enters the agricultural by-product wholesale market
Inner Mongolia branch
Founded in July 2018
Huo and Hot City will return to Hai Liang Square, West Central Mount, in the civilian area.
The service covers the entire country, and the response is right there.
Diverse cooperative biosphere
Building open and win-win industrial ecology and working with excellent partners to promote innovation in the AI data industry
Data group
Close collaboration between the data layer/trading grounds/hospitals to contribute to market prosperity of health data sets
Medical Institutions
Doctors are involved in data collection, cleaning, labelling, etc. to advance the development of paramedical megamodels
Pharmaceuticals and AI Enterprises
Customized data services meet individualized needs and work together to promote innovative development in the health sector
Research Institutes
Together with universities such as Central South University and Hunan University to build joint research bases in productive sciences to sustain innovation
Full WorkflowCompliance · Data security
Strict compliance with domestic and foreign data security regulations and development of a system of traceability-based compliance data services
Regulatory alignment
Align with China's Personal Information Protection Law, Data Security Law, and project agreements across collection, processing, and delivery (we do not claim HIPAA certification).
End-to-end de-identification
Implement de-identification processing from the data collection end, remove direct identifiers such as names and ID numbers, and virtualize sensitive information such as rare diseases to ensure that data can not be traced back to personal identity.
The authorization chain is traceable.
Establish a complete data authorization link traceability mechanism, each dataset can trace its data source, authorization chain and processing records to meet compliance audit and regulatory requirements.
Steady ahead, steady development
15 years of focus on the AI data service, and every step of the way is witnessing the flourishing of China's AI industry.
Establishment
Changsha Langhui Information Technology Co., Ltd. was formally established to start the journey of data service and establish a development strategy with technological innovation as the core driving force.
Tech Focus
Obtained software enterprise certification, launched the LangData data processing platform, established campus-enterprise cooperation with Central South University and Hunan University, and laid the foundation for industry-university-research.
Professional accreditation
Through ISO9001 quality system certification and CMMI level 3 international certification, it has accumulated more than 100 software copyrights, and its comprehensive strength has been recognized by the industry.
Strategic Upgrade
It has been certified as a high-tech enterprise, Huawei Kunpeng technology certification, and class B qualification for the integration of classified information systems, and has obtained the double certification of provincial-level and municipal-level contract-abiding and trustworthy enterprises.
Frontline AI Layout
Comprehensively upgraded to a cutting-edge AI data infrastructure provider, laying out six cutting-edge AI capabilities, serving the world's outstanding AI companies, and promoting the innovation and development of the industry.
Professional qualification
15 years of technology accumulation · 9 + core qualification categories · 176 software copyrights
High-tech Enterprise
GR202343000956 · 2023.10.16
CCMI3 Level Certification
International software maturity assessment
Software Enterprise Certificate
2023-2024 . Eight consecutive years of certification
ISO 9001 Quality Management
2020.7.9-2021.7.9
176 software rights
Cover areas such as land demolitions, smart cities, environmental monitoring, emergency management, property management, squatter tickets, educational systems, etc.
School Cooperation and Production Studies
Together, we'll build an integrated and innovative culture of production.
Central South University
Talent development baseProfessional development and project development base, in-depth research and development
Hunan University
Industry-academia baseThe Institute of Design Arts Co-operation in Research and Development in Innovation Design
Changsha University
Employment internshipsStudent Employment Internship Practice Base, School Co-Facilitators
The results of the integration of industrial studies are significant
With Langhui,Intelligence for the Future
You are a government agency, a medical institution or a technology firm.
We can provide you with high-quality data service solutions.