Exclusive Interview with Professor Nie Zaiqing: Is the Biopharma Version of "AI Lobster" Here? Shuimu Molecular's Dual-Wheel Model of "Proprietary Pipeline + Tool Empowerment" | Tong's Family

Background

Shuimu Molecular, a Dalton Family company incubated by Tsinghua University's Institute for AI Industry Research (AIR), has emerged as one of the most promising benchmark enterprises in AI-driven drug discovery. Its core achievements include the ChatDD intelligent agent framework, the OpenBioMed open-source platform, the Shuimu Zhiliao consumer-facing "super brain" tool for life sciences, and multimodal large language models — all supporting a dual-wheel drive strategy of "tool empowerment + proprietary pipeline R&D."

The following article is republished from VBData.

The deep integration of artificial intelligence and biomedicine is reshaping the fundamental innovation logic of the global pharmaceutical industry.

According to iiMedia Research, the global AI pharmaceutical market is projected to reach $2.99 billion by 2026. China's market is growing even faster — expanding more than sevenfold between 2020 and 2025.

This market heat is not without cause. At the J.P. Morgan Healthcare Conference in early 2026, multinational pharmaceutical firms and tech giants like NVIDIA announced major partnerships one after another. Domestically, leading pharmaceutical companies are actively pursuing BD deals and global expansion, with AI technology becoming another calling card for China's biomedical sector overseas, following biologics.

01

Tsinghua AIR Incubation + Former Alibaba AI Leader at the Helm

Professor Nie Zaiqing, GDS Professor at Tsinghua University and Chief Researcher at AIR, serves as the company's Chief Scientist

Shuimu Molecular's origins lie in the research soil of Tsinghua University's Institute for AI Industry Research (AIR) and the cross-disciplinary accumulation of its core team.

Founded in June 2023 and incubated by AIR — established by Academician Zhang Yaqin — Shuimu Molecular is dedicated to building foundational large models and intelligent agent platforms for the life sciences. The company's central figure, founder and Chief Scientist Professor Nie Zaiqing, is a GDS Professor at Tsinghua University, Chief Researcher at AIR, and the principal investigator for the new drug R&D large model direction under the National Key R&D Program's BTIT integration project during the 14th Five-Year Plan period.

Before joining Tsinghua, Professor Nie was a Principal Researcher at Microsoft Research Asia, where he led technical development for Microsoft Academic Search and the Microsoft natural language understanding platform LUIS. He later joined Alibaba as head of the AI Lab in Beijing and Chief Scientist for Tmall Genie, spearheading the strategic smart assistant project.

In 2020, Professor Nie joined AIR and redirected his industry focus toward biomedicine. "We judged at the time that the technological transformation brought by this wave of large models would be enormous — no less impactful than any industrial revolution we've experienced. Every industry will be rewritten by large models and intelligent agent technology, and biomedicine is the best entry point."

Professor Nie explains: "This industry has vast amounts of unique data, especially protein data, with its own 'language' and grammar — amino acid sequences determine protein structure and function. This uniqueness gives us the opportunity to build multimodal large models and products that are distinct from general-purpose models and truly solve industry pain points, achieving cross-modal translation between natural language and protein language, and enabling AI to genuinely serve biomedical research innovation."

Based on this assessment, Shuimu Molecular chose to enter the AI pharmaceutical track with multimodal understanding of "protein language + natural language" at its core. Today, the company has assembled a world-class team spanning research, engineering, and commercialization — with Academician Zhang Yaqin and Professor Ma Weiying as co-founders, former WuXi AppTec VP Xing Jie as Chief Operating Officer, and former Pathos Senior VP Dr. Pan Qi as Chief Scientific Officer of Biopharmaceuticals — forming a dual-core advantage of "AI technology + drug R&D experience."

02

Shuimu Molecular Is Becoming an AI Large Model-Empowered Drug Discovery Innovator

With the addition of Dr. Pan Qi — a veteran international biopharmaceutical R&D and management expert with over two decades of end-to-end experience spanning target discovery, drug development, clinical translation, and strategic partnerships — as Chief Scientific Officer of Biopharmaceuticals, Shuimu Molecular has gained decisive momentum. She oversees the company's biopharmaceutical and ChatDD (AI agent-driven human-machine collaborative drug discovery) research strategy, R&D system development, AI drug pipeline advancement, and global partnership strategy, while building teams to achieve wet-dry lab closed loops that inject core momentum into the company's long-term competitiveness and sustainable growth.

Shuimu Molecular has established its first proprietary pipeline in the autoimmune disease space. The chosen indication is a class of autoimmune diseases regulated by complex multi-signal pathway networks, where existing therapies targeting single pathways have limited efficacy. Based on systematic pathogenic mechanism analysis by ChatDD-Insight, the company decided to use AI to design bispecific antibodies that simultaneously block two core signaling pathways. Computational design via the ChatDD-Discovery platform has been completed, yielding multiple high-performing candidate molecules, with wet lab and preclinical validation to follow.

While advancing its pipeline, Shuimu Molecular is simultaneously building two core capabilities: an AI-driven bispecific antibody design platform, and a dual-target combination insight and validation system for autoimmune diseases.

Going forward, Shuimu Molecular will extend these capabilities to oncology and additional therapeutic areas, using AI to systematically and rapidly build a portfolio of innovative drug pipelines.

03

Dual-Product Breakthrough: OpenBioMed Open Source Disruption, Large Model Product Shuimu Zhiliao Democratizing the Industry

Under its core team's leadership, Shuimu Molecular has achieved rapid development since founding, with milestone results emerging continuously. The company has launched the world's first biomedical R&D intelligent agent ChatDD, the first multimodal open-source large model BioMedGPT, and the commercial large model ChatDDFM®. According to company disclosures, Shuimu Molecular has partnered with 30+ well-known brands, with 2025 full-year order value growing over 100% year-on-year and a prospective client base exceeding 100 organizations.

This momentum continues into 2026. Recently, Shuimu Molecular's product matrix achieved key upgrades: the B2B open-source platform OpenBioMed Skills was officially released, and the industry large model product Shuimu Zhiliao® for consumers entered public beta. This "dual empowerment" strategy marks Shuimu Molecular's formation of a "B2B empowering professional R&D, consumer-facing democratizing industry access" dual-product matrix, as the company transitions from technology validation to ecosystem building.

■ OpenBioMed Skills: The "AI Lobster" Breakout App for Biomedicine

"If we want to create the biomedical version of 'crayfish,' what's most lacking is biomedical professional skills." OpenBioMed Skills is the product of this thinking. Its underlying logic stems from Professor Nie Zaiqing's team's deep insight into pharmaceutical R&D pain points: "In traditional drug research, many steps rely solely on veteran expert experience, while general skill libraries mostly just wrap tool documentation without truly solving complex scientific problems."

OpenBioMed Skills' core breakthrough lies in systematically capturing knowledge scattered across academic journals, tool documentation, and experts' minds into a shareable, iterable skill library — enabling researchers without deep programming backgrounds to build and run end-to-end pharmaceutical workflows through natural language conversation. The initial release of 45 curated skills spans 5 categories, covering biochemistry/drug discovery, protein analysis and design, single-cell omics analysis, plus auxiliary skills for data retrieval and tool invocation.

Skills architecture diagram, courtesy of Shuimu Molecular

Compared to general-purpose skill libraries on the market, OpenBioMed Skills offers significant differentiation. It focuses on long-form workflow design, enabling multi-tool collaboration, end-to-end execution, and real-time feedback adjustment. It introduces human-machine collaboration, integrating expert knowledge and decision-making into AI skills to create "digital twins of expert brains," with built-in domain interpretation frameworks for contextualized professional data analysis — "like the biomedical version of OpenClaw." Additionally, the product supports fully localized deployment, with all data processing and model inference completable in users' private environments, fundamentally preventing data leakage.

Professor Nie Zaiqing notes, "The launch of OpenBioMed Skills marks the paradigm shift of biomedical AI from 'tools available' to 'expert capabilities reusable.'"

Currently, OpenBioMed Skills has been deployed on the OpenBioMed platform, with direct access to proprietary multimodal models including PharMolixFM, BioMedGPT-R1, and MutaPLM. Code is open-sourced on GitHub, and experts can remotely direct biomedical agents via Lark, DingTalk, and WeChat.

■ "AI Life Science Intelligence Assistant" Shuimu Zhiliao

Alongside its B2B open-source product, Shuimu Molecular has launched its consumer-facing product "Shuimu Zhiliao" in public beta. Derived from the internal ChatDD-Insight model, it focuses on intelligence interpretation and knowledge democratization in life sciences, filling the market gap for professional life science AI tools on the consumer side.

The life science research field generates massive amounts of intelligence daily — new technology platforms, AI models, targets, preclinical drug development progress, clinical trial updates, medical devices, biomanufacturing, pipeline transactions, financing information, and more. Shuimu Zhiliao works 24/7, personally filtering out repetitive, generic noise for each user and delivering only genuinely high-value content. No more needle-in-a-haystack searching — information screening efficiency improves 10x.

Meanwhile, based on the ChatDDFM® model family, Shuimu Zhiliao provides tiered services matched to users' knowledge levels: plain-language explanations of professional concepts for industry newcomers; deep yet accessible interpretations for experienced researchers; and authoritative knowledge graphs with support for in-depth exchange and AI consultation for expert-level users.

"Shuimu Zhiliao is more of an 'intelligence assistant,'" Professor Nie Zaiqing introduces. "It can interpret papers published today, industry developments, and answer specific user questions. It translates the esoteric jargon, industry terminology, and complex mechanisms of drug discovery into accessible language. We've set up deep thinking mode, auto-decision mode, and 'Xiaohongshu' style mode — hoping to both help researchers improve information acquisition efficiency and promote biomedical knowledge democratization, achieving the 'equalization' of knowledge that previously belonged only to veteran alchemists."

Additionally, the product is currently completely free, though paid tiers may be considered for future scenarios like deep reports and professional consultation that consume substantial GPU resources. B2B clients may also opt for internal deployment of a professional version of Shuimu Zhiliao, further expanding product coverage.

Based on these products, Shuimu Molecular has established deep partnerships with dozens of leading biomedical enterprises, hospitals, and research institutions, forming a diversified business model of "technology output + collaborative R&D."

On global expansion, Shuimu Molecular has initiated relevant planning across two paths: tool export and pipeline export. Partnerships have been reached in the tool export domain, with initial progress in East Asian markets, and gradual global expansion to follow.

When AI technology truly becomes a general tool for scientific researchers, the singularity moment for biomedical research may accelerate. Shuimu Molecular's exploration is a microcosm of this transformation — not replacing experts, but enabling their experience and intuition to be scaled through intelligent agents; not working in isolation, but making cutting-edge AI capabilities industry infrastructure through open-source ecosystems. From Tsinghua laboratory to industry frontlines, from open-sourcing code to building proprietary pipelines, Shuimu Molecular's story continues.

END

ID: daltonventure

Long press to follow

Recommended Reading