Amid the many industry giants on stage at WAIC 2026, Matwings Technology's self-developed MatwingsVenus™ (Xiaowu™) platform stood out to win the "Signature Exhibition Treasure" award, and its project lead, Gen Z PhD student Tan Yang, drew extensive coverage from mainstream media during the event.

The following is a full reprint of a special feature from Youth Daily (《青年报》), documenting the story of a young researcher who entered the AI+biology crossover field driven by a family member's illness, leading a team with an average age in the early twenties to overcome difficult challenges, and sharing the innovative practice and youthful research commitment behind Chinese AI-enabled life science technology.

Reprinted from Youth Daily with authorization 

Reporter: Zhu Bin 

Original article (in Chinese):

 http://www.why.com.cn/wx/article/2026/07/17/17842888641296148616.html

On July 17, the 2026 World Artificial Intelligence Conference (WAIC) officially opened, bringing together a host of top AI industry giants in Shanghai.

On a stage contested by industry leaders including Alibaba and Baidu, a homegrown AI product rooted in biological research stood out, winning the conference's "Signature Exhibition Treasure" award alongside the industry giants.

Surprisingly, this year's only AI for Science (AI4S) product selected for the honor was not developed by a team of veteran industry professionals, but by Gen Z scientist Tan Yang and his young team. Behind the young researcher's extraordinary achievement lies a story of channeling technical skill toward a compassionate purpose: healing illness through code.

01

A single agent that compresses R&D cycles by tenfold

"No deep coding required, no complex model building needed — issue instructions just like everyday conversation, and the AI can independently complete protein design, prediction, and end-to-end validation." Taking the stage at a world-class AI event for the first time, Gen Z PhD student Tan Yang confidently introduced his team's innovative creation: the MatwingsVenus™ (Xiaowu™) AI protein design platform.

This is a research platform that genuinely delivers "R&D through conversation." It not only runs on a self-developed AI agent but also connects the entire intelligent, automated research loop through lights-out laboratory technology, linking the complete workflow from virtual design proposals to real-world experimental execution. As team lead Tan Yang put it: "Simply put, it lets AI 'think through solutions' on the computer, and then actually run the experiments in the physical world."

The platform breaks through long-standing pain points in traditional protein R&D. He pointed out: "For a long time, protein R&D has relied heavily on researchers' accumulated experience. Countless researchers engage in trial and error in the lab, with high costs, lengthy development cycles, and success rates of less than 1%. Multi-year development timelines, tens of thousands of ineffective experiments, and extremely low translation rates have been three major mountains constraining progress in biological research. Our self-developed AI platform has rewritten this industry reality."

The platform compresses development cycles from 2–5 years to just 2–6 months, reduces the number of required experiments from tens of thousands to around a hundred, and lifts R&D success rates to 30%. Today, the technology has taken root, achieving deep collaborations with over 40 enterprises and enabling industrial applications in key fields including environmental protection and healthcare, allowing scientific breakthroughs to truly benefit society.

In the conference exhibition area, a fresh bottle of recycled plastic pellets offered a direct demonstration of the technology's broader social value. In nature, plastic-degrading enzymes can take hundreds or even thousands of years to break down waste plastic; the team's AI protein design model improved and upgraded natural degrading enzymes to dramatically accelerate plastic degradation efficiency, rapidly converting waste plastic into recycled raw materials.

02

Code carrying the warmth of healing family illness

"Human intelligence alone struggles to break through cognitive limits to create entirely new substances, but AI can." This is the original motivation driving Tan Yang's deep focus in the field. His aspiration is to lower the threshold for scientific innovation through AI, enabling research teams and SMEs everywhere to create protein products that benefit more people.

Born in 2000, Tan Yang grew up during the era of the internet's explosive growth, developing an early connection with computers and exploring possibilities through code and programming. "When bored, I would modify an Android phone system into a Windows system" was a regular part of his life. After leading a team to the FIRST Tech Challenge (FTC) World Championship during secondary school, he chose to deepen his focus in computer science. For the college entrance exam, he resolutely applied for software engineering, eventually becoming a joint doctoral student in computer science at Shanghai Jiao Tong University and the Shanghai Chuangzhi Institute — steadily advancing along the digital technology track.

Although the computer science path offered broad prospects, Tan Yang deliberately chose the more difficult and contested "AI + biology" crossover field. Behind this unusual decision lay a young man's deeply personal conviction. His grandfather and uncle had long struggled with diabetes, and his grandmother had been diagnosed with Alzheimer's disease — despite early detection and active treatment, none could be cured. Watching his loved ones suffer from illness left him with a profound sense of helplessness.

To explore the possibility of a cure, he immersed himself in vast amounts of literature on Alzheimer's disease R&D, learning about the latest research frontiers. In countless late nights, he began to reflect that the significance of AI might extend beyond algorithms and models — it could become a genuine aid in treating human disease. "I've studied computer science for over a decade. Can I use my own expertise to do something for my family, and for the many others suffering from illness?" Carrying this simple wish, he stepped out of his comfort zone and began a cross-disciplinary research journey. Since Alzheimer's treatment requires monoclonal antibody drugs, he zeroed in on protein R&D, spending nearly two years refining an early prototype of a protein design agent — the Venus large model — and using it to successfully boost the activity of the Cas12i3 protein fivefold.

"The benefit of crossing disciplines is being able to see landscapes invisible from any single field," Tan Yang reflected. When computational thinking meets the complexity of biology, new possibilities emerge. His work quickly drew attention from the industry — many researchers, upon seeing the technical prototype of Venus in the open-source community, proactively reached out by email to discuss and exchange ideas. Recognition from frontline researchers further convinced the young PhD student that this path was worth pursuing.

03

A team of Gen Z researchers writing a new answer sheet for scientific research

During the bottleneck transition from research to project implementation, Tan Yang met his mentor, Professor Hong Liang. As founder of Shanghai Matwings Technology and Distinguished Professor at Shanghai Jiao Tong University, Hong Liang comes from an interdisciplinary physics and computer science background — which gives him a notably different perspective on protein design problems than traditional biologists. After a single conversation, the two found their thinking highly aligned: use AI to redefine the approach to protein R&D. They decided to make it happen together, throwing their full effort into advancing the protein AI agent project toward real-world implementation.

"More than 70% of our company's employees were born in the 1990s, and the average age of our R&D team is even younger," Hong Liang said. Building on the foundation model, this group of young researchers worked day and night, refining code line by line to build the entire MatwingsVenus™ platform from scratch.

Compared with industry giant Google's AlphaFold, Tan Yang's platform delivers end-to-end design directly from sequence to function, aligning more closely with real-world industry needs and covering the entire chain of protein design, validation, and experimental generation.

"Building an agent today isn't the hard part," Tan Yang admitted candidly. The real challenge is real-world implementation. "Computational overload, user privacy protection, real-world scenario adaptation — countless engineering challenges must be tackled one by one through repeated debugging."

In April of this year, the AI protein design platform was officially launched. Yet neither technological success nor accumulated honors has changed Tan Yang's original mission. To this day, the platform remains freely accessible to the whole of society. "I want to build it into public infrastructure for life science research for all of humanity," the young Gen Z researcher said with firm conviction.

Biography:

Tan Yang

Born in 2000, Research Scientist at Matwings Technology AI Lab, lead developer of the dialogue-driven protein R&D agent MatwingsVenus™ (Xiaowu™). He is currently a 2025 cohort joint PhD candidate at Shanghai Jiao Tong University and the Shanghai Chuangzhi Institute.

He was selected for the 2025 China Association for Science and Technology (CAST) Young Talent Doctoral Training Program, and has received more than 20 provincial/ministerial-level and above awards including multiple National Scholarships. He has published 13 papers as first or co-first author at top conferences and journals including NeurIPS, ICLR, and eLife, with 26 total publications and over 600 citations. He serves as a reviewer for venues including Nature Machine Intelligence, ICML, and ICLR. His research outputs rank first on Harvard Medical School's ProteinGym protein engineering benchmark; his open-source models and datasets have accumulated over 300,000 downloads on Hugging Face. He also contributed to development of Shanghai Jiao Tong University's Venus series large models and the Zhaoyan large model of the Shanghai-Chongqing Artificial Intelligence Research Institute.