「智源深澜」获天使轮融资，构建数据驱动的AI生物分子设计平台

Core Insights - The article discusses the recent angel round financing of "Zhiyuan Shenlan," which raised several million yuan, led by Woyan Capital, with participation from Tianfeng Capital and other angel investors. The funds will be used for building a generative AI platform for biomolecules and a self-driving molecular function evolution platform, as well as for market expansion [3][4]. Group 1: Company Overview - Zhiyuan Shenlan was established in 2024, incubated by Megia Technology, focusing on data-driven biomolecule design and manufacturing. The founder, Dr. Wang Chengzhi, has over 20 years of experience in the life sciences field [3][4]. - The company aims to leverage generative AI to transform the research paradigm in life sciences from "large-scale trial and error" to "precise design and creation" [3][5]. Group 2: Technological Advancements - The emergence of AlphaFold 2 has predicted over 200 million protein structures, covering most known proteins on Earth, but the industry is more focused on protein functions rather than just structures [3][4]. - Zhiyuan Shenlan is optimizing for "function" by exploring functional needs in practical applications, utilizing a self-driving automated experimental platform to efficiently generate functional data [4][5]. Group 3: Future Vision - The company is working towards a data-driven bioengineering and molecular design platform, evolving AI for Science from a 2.0 "navigation design engine" to a 3.0 "scientific intelligent autonomous platform" [5][6]. - In the future, AI is expected to autonomously design, execute, and iterate entire research experiment loops, with human scientists focusing on key questions and strategic directions [5][6]. Group 4: Key Breakthroughs - Three key breakthroughs are necessary for the autonomous intelligent evolution platform: a unified coordinate system for AI understanding, an autonomous decision-making AI agent, and an automated intelligent experimental platform [6]. - Zhiyuan Shenlan proposes a "ten-step" roadmap for AI4S 3.0, aiming to learn existing human knowledge, propose hypotheses, validate experiments, and ultimately make scientific discoveries that surpass human intuition across multiple scientific fields [6].