Originally reported by AIThority.com. Original publication date: 7 April 2020.
Overview
Tencent Holdings Limited announced a deepened collaboration with the World Health Organization (WHO) to provide technology support during the COVID-19 pandemic. As part of that effort, Tencent open-sourced a COVID-19 self-triage assistant and made it available on GitHub for developers around the world. The tool enables preliminary self-evaluation for potential coronavirus infection and delivers prevention tips in plain language.
This release followed an earlier open-source contribution: a COVID-19 live updates module (TH_COVID19_National) that Tencent published the previous week. According to Tencent, that module had already answered approximately six billion pandemic-related queries in China over a two-month period.
What the Tool Does
Tencent Health built the self-triage assistant to help individuals carry out a structured self-check for potential COVID-19 infection, with the stated goal of reducing unnecessary hospital visits and the associated risk of cross-infection. The assistant translates professional medical guidance into easy-to-understand messages suitable for the general public.
Key capabilities include:
- Symptom intake via conversational interface: The tool conducts rounds of virtual dialogue to collect information such as contact history with confirmed COVID-19 patients, current symptoms (e.g., fever, cough), and pre-existing health conditions.
- Risk assessment: Based on the gathered data and professional medical guidance, the assistant estimates the likelihood of infection.
- Preliminary recommendations: Users receive suggestions on whether to seek in-person medical care or self-quarantine.
- Shareable report: After the self-evaluation, users receive a summary report they can bring to a doctor for further consultation.
Underlying Technology
Tencent Health applied natural language processing (NLP) for medical language to build a knowledge graph of respiratory disease diagnosis information. This graph powers the question-and-answer flow within the assistant, allowing it to map user-reported symptoms and history to structured diagnostic pathways.
The tool follows the Diagnosis and Treatment Protocol for Novel Coronavirus Pneumonia (Provisional 7th Edition) published by China’s National Health Commission on 4 March 2020, and was reviewed and approved by the Chinese Preventive Medicine Association.
Open-Source Structure and GitHub Availability
The COVID-19 self-triage assistant is released as a companion to TH_COVID19_International, Tencent Health’s existing open-source COVID-19 international live updates module on GitHub. Together, they form Tencent’s primary open-source contribution to the global pandemic response.
For developers, the open-source code offers a practical starting point:
- Adapt the diagnostic logic to reflect guidelines published by local health authorities in different countries or regions.
- Integrate the triage flow into existing health applications, hospital portals, or government services.
- Localise the conversational interface for different languages and cultural contexts.
Adoption and Scale
At the time of publication, the tool had already been deployed to support:
- 40 disease control and prevention centres in China
- More than 300 hospitals
- Over 30 independent software vendor (ISV) partners
- Deployments across 17 provinces
- Tens of millions of end users
Tencent’s Broader Open-Source Footprint
As of March 2020, Tencent had open-sourced 98 projects spanning cloud computing, big data, AI healthcare, and network security. The company’s GitHub presence had accumulated more than 287,000 stars across its repositories, placing it in the top 10 of GitHub’s global open-source star rankings according to third-party data cited by Tencent.
Relevance for DevOps and Engineering Teams
For engineering and DevOps teams involved in building health-tech applications or public information services, this open-source release provides a reference architecture for conversational triage systems built on NLP-driven knowledge graphs. Teams looking to deploy similar tools should evaluate:
- How to version-control and update clinical guideline data as protocols change.
- Compliance with local medical device software regulations before deploying in a clinical context.
- Localisation pipelines for language and regional health authority guideline integration.
- Logging and audit requirements when handling sensitive health-related user data.
The source code for both the self-triage assistant and the live updates module was made available on GitHub. Developers should search for Tencent’s official GitHub organisation (Tencent) and look for the TH_COVID19 repositories to find the most current versions.