Here are 14 startups that haven't been covered yet.

"Always celebrate the risk-takers." By Muxin Xu and Jiaxiang Shi

Edited by Zhiyan Chen

Today, self-promotion by startups has become utterly commonplace. From WeChat official accounts, Red Xiao's platform, Jike, Xiaoyuzhou, to Bilibili, founders meticulously craft personal brands, sharing their growth stories, funding wins, and product methodologies. This wave of communication coursing through the business world manufactures an illusion that "everything can be witnessed."

Yet this highly "curated" and "visible" entrepreneurial practice may represent only the tip of the iceberg.

Beneath the surface lies a far vaster "underwater" world. There are countless early-stage projects that have never been reported on in detail, that have raised just one round of funding, or that haven't raised any funding at all. They fall outside mainstream narratives. The reason is obvious: whether traditional media or self-media, the decision to cover a company or founder hinges on the traffic and attention they can generate. Data has long become a crucial metric for media commercialization.

We're no exception.

But through our market contacts and conversations, we've still learned about some of these underwater projects from investors, financial advisors, and other channels. In speaking with these founders beneath the surface, we quickly discovered that while capital and traffic direct the vast majority of attention toward headline projects, far more people are quietly exploring in narrow niches.

This is the starting point for this piece from Undercurrent Waves ("暗涌Waves"). We hope it can provide the investment industry with some valuable firsthand information. And we hope even more that it's just a beginning. We invite entrepreneurs in the 0-to-1 phase to talk with us, to get their stories out into the market. In this transitional era, at a moment when most people are pursuing safety and certainty, there are still those willing to throw themselves into vast uncertainty — and that deserves to be cherished and seen.

After all, regardless of outcome, adventure and innovation are inseparable.

Entrepreneurs seeking coverage can reach us at: chenzhiyan@36kr.com.

01 Kira.art

· One-liner: Kira.art is an AI-powered photo editing and sharing community that lets users retouch, filter, and express visual creativity in a more natural way.

· Company overview: Kira.art is built on the vision of "letting everyone easily create and share beautiful images," blending AI image editing with social sharing. Users can describe desired effects in natural language, or use tools like Replace BG, Inpaint, Expand, Redux, and Eraser for precise adjustments. With "contextual continuous creation" as its core design philosophy, every edit flows naturally within the same context, making the retouching process smoother and more coherent. At Kira, the boundary between editing and sharing dissolves; the creative process becomes light and intuitive. Kira currently supports Replace BG, Inpaint, Expand, Redux, Eraser, and other image editing functions, with "contextual continuous creation" as its central design principle, allowing each modification to naturally continue within the same context.

· Future development: Going forward, Kira will evolve from an intelligent retouching tool into a complete AI image community. Users will be able to generate, edit, publish, discover, and interact on the platform, making AI a new medium for visual expression and social connection. Kira's goal is to build a visual ecosystem co-driven by AI and users, where every image can be created, shared, and reinterpreted.

· Founding team: Founder Longbo Chen is an early ByteDance employee and TikTok's first engineer. Other team members come from ByteDance and top universities in the UK and US, bringing combined backgrounds in product, algorithms, and visual design.

· Comparable brand (company): AI Instagram.

· Business model: Kira uses a freemium model; users can access basic features for free.

· Funding status: Has raised several million USD.

02 WetuTalk

· One-liner: A relaxed and efficient AI-powered spoken language teaching app.

· Company overview: Founded in January 2025, the founding team comes from top-tier internet companies with combined experience in online education and AI applications. The product targets adult spoken language learners, began small-scale beta testing in May 2025, and will launch its official app in November 2025. Current team: 8 people.

· Future development: Starting from China's adult English market, expanding within two years to users in East Asia, North America, and Europe, with support for multiple languages. The ultimate goal is to become a vertical spoken language learning platform integrating teaching, practice, and communication.

· Founding team:

• CEO & Growth Lead: Steven, with years of online education experience, formerly senior growth manager at ByteDance Education and VIPKID; among the first wave of AI application practitioners, formerly growth lead at a top domestic AI application company; also led growth strategy at Tencent and Meituan.
• Product Lead: Hope, with years of online education experience, formerly product lead at ByteDance Education; among the first wave of AI application practitioners, community product lead at a top domestic AI application company; also served as product lead at Tencent and Baidu.
• Engineering Lead: Levi, formerly backend lead for ByteDance's international payments, responsible for system design and development handling tens of millions of daily orders, with extensive hands-on experience in high concurrency, stability assurance, and monitoring infrastructure.
• Algorithm Lead: Lex, formerly lead algorithm engineer for large models at a military-industrial enterprise, contributor to an open-source project with 36k GitHub stars, with extensive experience in model training, fine-tuning, and prompt engineering.

· Comparable brand (company): Speak.

· Business model: Basic teaching features are free; premium features require subscription.

· Funding status: Completed angel round in August 2025, invested by Changlei Capital.

03 MosuMosu

· One-liner: A personalized oshikatsu information platform built for the subculture world, where AI makes oshikatsu incredibly easy and incredibly safe based on your attributes.

· Company overview: Founded in February 2025, the founding team comes from top BAT companies with combined experience in Gen Z content creation ecosystems and AI consumer applications. The product focuses on the Japanese market, deeply rooted in youth culture and female-oriented content. Current team: 6 full-time members.

· Future development: A Weverse + SmartNews for oshikatsu, exclusively for the next generation of women.

· Founding team/Core team:

• Jayde: Nanyang Technological University, College of Computing and Data Science. Former BAT market strategy analyst & PM, serial entrepreneur.
• Other team members formerly worked at ByteDance, Red Xiao's platform, and the "AI Six Little Dragons."
• Beyond our professional labels, we're independently filming oshikatsu documentary films in both Chinese and Japanese.

· Comparable brand (company): Reddit / Feedly / Flipboard & Weverse.

· Business model: MosuMosu adopts a "brewing" model rather than simply "collecting tolls."

• MosuMosu offers AI-driven membership subscriptions;
• Event ticketing revenue share, brand partnership advertising, and merchandise e-commerce referrals, providing fans with a one-stop应援 (support) experience.

· Funding status: Angel round.

04 Gelo (阁楼)

· One-liner: A company providing AI-powered psychological counseling services, most recently launching AI psychological advisor "Ada."

· Company overview: "Gelo" was founded in 2021 as an online platform centered on professional psychological counseling. It uses a standardized system (not simply a marketplace matching users with counselors), functioning more like an "online psychological services company" offering one-on-one video and text consultations. Their AI psychological advisor Ada evolved from the Copilot system accumulated over the company's first few years. It focuses on low-cost, accessible,碎片化 (fragmented-use) psychological companionship and consultation tools. By end of 2024, Gelo fully pivoted to consumer-facing AI productization — with AI directly providing psychological advisory services to users — while Ada is envisioned as the underlying infrastructure for psychological advisors.

· Future development: China's psychological counseling market is roughly in the hundreds of billions of RMB, but under traditional models, counselor quality varies wildly, prices are high, and coverage is low. AI介入 (AI intervention) actually reconstructs the supply side, because AI makes "supply standardization" and "cost reduction" in psychological services possible. AI's advantages lie in "low defensiveness + consistency + replicability" — users are more willing to open up to AI (low defense, strong sense of privacy and security). Going forward, Gelo hopes to make Ada into psychological AI infrastructure, embedded in various products, with overseas expansion into East Asian markets. In the long term, they hope AI can possess empathy and ethics, becoming "intelligent agents with heart."

· Founding team:

• Founder Qiuyang holds bachelor's and master's degrees in computer science from the University of Southern California (USC). In 2019, a life transition led him to psychological counseling, during which he discovered numerous industry problems and market pain points, prompting him to found Gelo. Qiuyang's entrepreneurial logic: start from user pain points → build standardized psychological counseling system → use AI to achieve accessibility and low-cost scaling.
• The other co-founder comes from a psychology background, responsible for professional model and consultation process construction.
• Current core team members come from three areas: psychology, AI algorithms, and internet product operations. The company already has mature Copilot systems, counselor standardization frameworks, and technical architecture.

· Comparable company: Ash (Slingshot), focused on AI chat-based psychological advisor combining CBT (Cognitive Behavioral Therapy) and DBT (Dialectical Behavior Therapy), which has raised $93 million. In Gelo's view, Ash operates in a high-awareness market where users understand the therapy concept, while China is a low-awareness market requiring more educational effort and closer alignment with user psychological cognition.

· Business model: Consumer side uses subscription model — for example, Ada is ¥40 for 220 messages, using low prices to acquire users and reduce the "trial cost" of psychological counseling, improving user penetration. B2B side goes through corporate EAP (Employee Assistance Programs): embedded in WeChat Work / Feishu as built-in psychological advisor. Gelo has already achieved break-even.

· Funding status: Has completed its Series A+ round. Previous investors include Legend Star, Ameba Capital, and FreeS Fund. Currently raising a Series B to fund consumer-side growth, international expansion, and continued product refinement.

05 KYMS AI

· One-liner: An AI-driven recruiting agent platform where multiple AI Agents collaborate to give everyone their own AI HR team.

· Company overview: Founder Bai Bing spent 18 years in the software industry, deeply involved in RPO (Recruitment Process Outsourcing) and BPO (Business Process Outsourcing). He observed that companies "want people, not resumes" — traditional human-dependent recruiting networks face dual constraints of capability boundaries and resource boundaries. As large language model technology matured, Bai saw the opportunity to restructure recruiting with AI Agents and launched his venture. KYMS AI helps companies break through both "capability boundaries" and "resource boundaries" in hiring, giving everyone their own AI HR team and automating the full process from demand analysis to talent delivery. The core product is an intelligent recruiting system composed of multiple AI Agents: a demand analysis Agent that understands job requirements and formulates recruiting strategy; a talent sourcing Agent that searches across multiple channels in parallel; an interview Agent that completes candidate assessment; and an intelligent screening Agent that enables precise matching and recommendation. Through bidirectional understanding of enterprise job requirements (B-side) and candidate capability profiles (C-side), the system achieves closed-loop automation from demand analysis, talent sourcing, intelligent screening, AI interviewing to result delivery. Since launching three months ago, KYMS.AI has served nearly 150 enterprises across 12 major industries, organized approximately 13,000 AI interviews, and accumulated 20,000+ candidate profiles.

· Future development: Restructure recruiting with AI to build "results-oriented" AI recruiting infrastructure. Near-term: complete Agent matrix construction and achieve full-process recruiting automation. Medium-term: build platform ecosystem enabling anyone to become a "Super HR" and monetize their capabilities. Long-term: accumulate tens of millions of job and talent data points, expand to more HR business scenarios, and become the capability-layer infrastructure of the AI era.

· Founding team:

• Founder Bai Bing, former Senior Vice President at Legend Software, built and managed outsourcing teams of over 1,000 people, serving 500+ global clients including Microsoft, Motorola, and ATOS with technology and talent services. Bai has been both the hiring demand side and the talent delivery side.

• Co-founding team includes serial internet entrepreneurs and HR domain experts, all Tsinghua EMBA classmates of the founder, forming a complementary team of "recruiting + AI + commercialization."

· Benchmark companies: Overseas benchmark is Juicebox (U.S.), whose product PeopleGPT is an AI agent-based recruiting matching platform providing RPO services through dynamic matching of demand and capabilities, focusing on "AI interview + automatic candidate delivery," currently widely adopted among U.S. tech companies and startups. Another reference is Mercor AI (U.S.), valued at $10 billion, which also provides AI-based recruiting matching and industry BPO services. Notably, China's recruiting market has larger demand volume, shorter cycles, and more complex processes — especially in "high-frequency hiring scenarios" like food service, retail, and tech. Therefore, the China market requires greater emphasis on "rigid demand + high frequency + delivery results" rather than generic AI resume matching.

· Business model: Core model is pay-per-hire — companies pay based on "successful hires" through AI Agents, not subscription fees. Average pricing: number of successful hires × commission rate. Revenue structure includes B-side payment based on hiring results (similar to RPO logic), plus C-side participation where candidates interview for free but their data enters the enterprise talent pool for subsequent reuse. Later extensions: enterprise subscription for long-term agent deployment + industry data value-added services + C-side intelligent job recommendation value-added services.

· Funding progress: Just started external fundraising, currently in angel round.

06 Zhidai Qianfan (智电千帆)

· One-liner: AI for Science — using AI to drive new materials R&D.

· Company overview: Zhidai Qianfan is an intelligent materials R&D solution provider combining self-developed algorithms, Hellmann-Feynman simulation (chemical computation models), and high-throughput experimental platforms. Target clients are advanced manufacturing enterprises (lithium batteries, polymers, daily chemicals, semiconductors, hydrogen energy, photovoltaics, etc.). Main outputs are improved R&D efficiency, shortened cycles, reduced experimental costs, and helping clients break through materials performance bottlenecks — forming a disruptive difference from traditional "experience + trial-and-error" experimental R&D models.

· Future development: Build a Chinese model of "AI for Science" — achieving a closed loop integrating AI, experimentation, and computing power.

· Founding team: Founder and CTO Yang Hao is a professor at Soochow University, Caltech PhD, previously participated in U.S. military research projects, with research focus on materials science and AI chemical computation. After returning to China in 2019, he focused on lithium batteries and accumulated rich interdisciplinary research achievements. CEO Jiang Xi and co-founder Cheng Tao are both Soochow University alumni. The company's core technology originated from the Soochow University professor team (Yang Hao, Cheng Tao), with years of deep cultivation in the AI + chemistry intersection. Over the past three to four years, the research group continuously undertook industrial client R&D contracts (such as Amperex Technology, Huawei Watt Lab, P&G). In late 2024, Soochow University issued an official document "supporting professors in entrepreneurship," which facilitated the team's independent registration of Zhidai Qianfan in June 2025 — the corporatization is a natural extension of "commercializing research achievements."

· Benchmark companies: Listed AI drug discovery leader XtalPi, AI drug R&D representative Insilico Medicine, domestic AI + materials R&D star project Deep Principle, and foreign computational chemistry veteran Schrödinger. Core differentiation: Zhidai Qianfan focuses on AI for Materials / Chemistry rather than AI drug discovery, and its technology combines self-developed algorithms with experimental apparatus ("hardware-software-algorithm closed loop") while competitors mostly remain at the software layer. The company has completed 10+ real delivery projects, with commercialization progress leading the industry by one to two rounds. Specific benchmark ambition: become "the materials R&D version of Schrödinger + XtalPi for China."

· Business model: Primarily B2B, targeting leading R&D departments in lithium batteries, chemical materials, daily chemicals, new energy, semiconductors, etc. Revenue from project-based service fees + algorithm/experimental deliverable payments. In 2025, locked confirmed orders of approximately RMB 20–30 million, with two-year framework agreements signed with multiple listed companies (including Amperex Technology, Huawei Watt Lab, P&G, Yanyi New Materials, NIO). Expected to break even in 2026.

· Funding progress: Currently in angel round, post-money valuation approximately RMB 100 million. 40% of funds for expanding ~2,000-card computing center; 60% for operations and cash flow reserve. Prefers long-term, rational capital.

07 Luvbotics (越伴动力)

· One-liner: Luvbotics focuses on developing innovative home companion robots, dedicated to creating an entirely new living experience.

· Company overview: Founded in September 2025, engaged in R&D, production, sales, and technical services of robots and related software/hardware products, covering home companion robots, humanoid robots, and ecosystem value-added applications.

· Future development: Luvbotics will continue to create safer, more lifelike robot products, shaping them as "scenario experts" and "emotional centers" for families.

· Team composition: Core team mainly comprises master's and PhD graduates from renowned institutions including BIT, HIT, Tsinghua, UCSD, and TU Delft, with product and R&D capabilities covering joint modules, robot bodies, embodied interaction, and decision-making.

· Benchmark companies: Groove X, 1x Tech, Figure AI.

· Business model: Consumer robot product sales, subscription services, and ecosystem value-added services.

· Funding status: Received ZhenFund seed round investment.

08 PaaaWOW

· One-liner: PaaaWOW smart pet collar — a pet training and lifestyle community that brings people and pets closer together.

· Company overview: Founded in 2024, already launched Kickstarter crowdfunding. Core products are the world's first fashion smart pet collar, and the world's first multimodal AI-based pet health and emotion collar.

· Future development: Plans to sell across platforms by mid-2026.

· Founding team:

· Duan Yuhao, CEO, from Beijing Academy of Artificial Intelligence;

· Yang Yang, CDO, from Porsche.

· Benchmark brands (companies): Fi, Tractive, Traini.

· Business model: Hardware sales, software value-added services.

· Funding status: 2024 seed round, 2025 angel round (Beijing Academy of Artificial Intelligence), valuation RMB 30 million.

09 Jiujian Technology (九间科技)

· One-liner: AI-driven deeply personalized learning platform.

· Company overview: Flagship product "Ke You Ji" (Class Travelogue) is an innovative AI learning product utilizing a powerful educational content generation engine and advanced concepts to provide efficient, personalized learning experiences. Intelligent AI twins tailor learning paths, content, and real-time feedback based on student progress and comprehension depth, achieving true individualized instruction.

· Future development: Plans to deepen AI educational content generation and interaction, learning analytics, expanding to more subjects, grade levels, and vocational skills training. Meanwhile, actively expanding internationally to build a cross-cultural, barrier-free learning ecosystem.

· Founding team: Li Chun, Peking University undergraduate, Columbia University AI master's, former senior engineer at Amazon and Google. Serial entrepreneur; co-founder & CTO at first startup XKool Technology, which grew to 300 people.

· Benchmark brands (companies): Duolingo, Khan Academy, Zebra AI.

· Business model: Subscription service.

· Individual users: Tiered monthly/annual subscription services unlocking full subjects, advanced AI instructor customization, deep learning reports, exclusive gamified content, and other features.

· Institutional users (B2B): Customized AI teaching solutions, teaching research platform integration, student management systems, and data analytics services for schools, educational institutions, and training centers — charged per user or service package.

· Value-added services: Advanced AI instructor/content customization: More advanced AI instructor avatars, specific topic course packages, or deep personalized content customization services.

· Online-offline integration: Partnerships with physical study rooms, summer camps, etc., providing AI-assisted teaching services, charging cooperation fees or commissions.

· IP cooperation: Collaboration with content creators and IP owners to AI-ize their content or character images for platform introduction, with revenue sharing.

· Funding status: Seed round.

10 Bian Xin (彼岸心智)

· One-liner: Provides solutions for smart wearable brands (e.g., imoo, Xiaomi).

· Company Overview: Founded in 2021 and based in Nanshan, Shenzhen. Its technology originates from the Institute of Psychology, Chinese Academy of Sciences. The company's core technology is an emotion recognition algorithm based on sympathetic and parasympathetic nervous system activity. Its main business is providing emotion recognition algorithms for smart hardware devices. Currently, the emotion features in leading domestic wearable brands' devices mostly run on Bian Xin Mental's algorithms.

· Future Plans: The company is currently preparing to take its smart hardware overseas, with plans for a consumer-grade emotion hardware product focused on mindfulness meditation and healing. Expected launch: January 2026.

· Founding Team: Sun Haiyang holds a master's degree in applied psychology from the Institute of Psychology, Chinese Academy of Sciences, and is a recognized expert in Shenzhen's science and technology expert database. He has years of experience in psychological equipment and affective computing. After graduating, he did marketing at a CAS-affiliated startup for three years before founding his own company.

· Benchmark: CyweeMotion.

· Business Model: Dual-engine drive of algorithm licensing and industry solutions. Currently B2B-focused, with consumer-grade smart hardware in planning.

· Funding Status: Angel round from CAS Angel Fund in 2023.

11 BDF

· One-liner: A Singapore-born essence skincare brand with global ambitions.

· Company Overview: A sustainable skincare brand from Singapore, currently centered on capsule-format products. It emphasizes zone-specific anti-aging and oil-based skincare. The brand owns its formulations and patents. Its capsules are exclusively produced by Brunei's Royal Pharmaceutical Factory to pharmaceutical-grade GMP standards; non-capsule products (such as water-based formulas) are manufactured at its own GMP facility in Singapore. Product design prioritizes emotional and sensory experience: macaron color palettes, Singaporean tin boxes, rattan weaving craftsmanship, spa-grade fragrance experiences, and content creation infused with Nanyang aesthetics.

· Business Model: Best described as a brand retail model of "DTC-first, offline distribution as supplement," built on in-house R&D and premium supply chains. Notably, the brand is considering a global store rollout and a diversified channel structure going forward.

· Future Plans: With Greater China as its primary market, it is gradually developing into a global brand with diversified presence in Southeast Asia (Vietnam, Indonesia, etc.) and Europe (France, etc.). Category-wise, leveraging its professional pharmaceutical factory system and global supply chain capabilities, it plans to extend into wellness and ingestible beauty — hoping to replicate FANCL's "health foods + skincare" model.

· Founding Team:

· Kris: Singapore-based partner and second-generation successor of the family business. Oversees Singapore, overseas markets, and factory management. His family has operated a cosmetics OEM factory in Singapore since 1979.

· Cathy: Leads comprehensive China operations including product development, launch, and marketing. In 2016, she co-founded Blé de Fonty with Jeff (a former L'Oréal R&D formulator with years of experience) and the current partner team.

· Benchmark Brands: Structurally, FANCL — admiring its "health foods + skincare" business model and its positioning around female beauty and wellness. In brand tone, Lush — hoping to make skincare fun, exploratory, and joyful rather than boring. As a category pioneer, Elizabeth Arden — Arden invented the capsule format, but BDF aims to build a "younger Arden" with packaging and product lines better suited to contemporary refined mothers and travel-oriented consumers.

· Funding Status and Needs: BDF has taken zero external funding over the past nine years, growing entirely on operating cash flow. It is now seeking outside investment for the first time, hoping to bring in investors with industry backgrounds — ideally those with Southeast Asian or European market operations and resources to aid international expansion, or those with positions and resources in health foods to help with category extension. Target raise: approximately $5 million.

12 Norman Walsh

· One-liner: A 65-year-old British footwear brand, now under new ownership and positioning itself as a global brand.

· Company Overview: Norman Walsh was founded around 1960, starting with track training shoes. It is a classic family business whose defining characteristic has been maintaining and operating its own factory in the UK for manufacturing. Its "Made in Britain" identity and handcrafted shoemaking are core assets. Beginning in 2018, Norman Walsh moved some production to China for China-only sales, but fell into operational difficulties during the pandemic. The brand's current steward, Charlie, had been its agent in China for 32 years, building the China business to roughly RMB 100 million in annual sales. At the end of 2023, Charlie completed a 100% acquisition of Norman Walsh using personal funds — acquiring both the brand and its British factory. The primary motivation was securing global ownership of the brand and eliminating the risk of being a major agent at the mercy of others.

· Future Plans: The top priority is reducing excessive revenue dependence on the China market and developing Norman Walsh into a truly global brand. Key focus markets: Europe, Japan, and the US. Product-wise, it plans to revive the brand's historical road running and trail/hiking shoe lines to create a powerful portfolio alongside its existing heritage collection, benchmarking against comprehensive sportswear brands like New Balance.

· Founding/Stewardship Team:

· Charlie, a serial entrepreneur. His first venture was founding a new footwear brand in the US, which was acquired by a Korean conglomerate in 2016.

· The current team is almost entirely international professionals, leveraging their backgrounds and networks to crack mainstream markets in Europe, America, and Japan. For example: the creative director is British, based in Tokyo, overseeing visual identity and product; the designer is German, formerly at PUMA; the development manager is a 30-year veteran of the Norman Walsh factory in the UK.

· Benchmark Brands: Primarily New Balance, as a successful comprehensive athletic footwear brand with product lines spanning heritage, running, and trail. Norman Walsh's planned revival of road running and trail lines is specifically intended to build a similar product matrix and compete in the same arenas. In niche segments, it also references specialist trail brands like Salomon and Hoka.

· Business Model: China market is primarily direct-to-consumer. Offline through brand-owned stores (already opened in Shanghai), online through owned e-commerce — currently its most mature and highest-revenue market. Overseas is primarily wholesale, selling through local distributors and retailers.

· Funding Status: As a family business prior to acquisition, Norman Walsh never raised external funding. As a wholly owned acquisition, the company currently has no outside investors. It is now casually seeking roughly $8 million in external funding. The ideal investor profile would be European family funds or deeply resourced industry investors with strong retail networks and connections in Europe, America, and Japan to support global expansion.

13 AILUO Technology

· One-liner: VIBE CREATING — an AI-driven vibe-making system that lets users interact with AI through natural language and "Lego-grade" modular assembly, enabling a from-zero-to-one physical hardware creation experience and ushering in an era where anyone can be a maker.

· Company Overview: Founded by a Harvard-MIT team, AILUO is an innovation company focused on AI + Creation Tech. With "AI as building blocks" as its core concept, it is dedicated to making AI a creative partner for humanity, helping users transform ideas in their minds into manufacturable physical objects through natural language description. Its flagship project, VIBE CREATING, is an integrated system combining an AI design engine + DIY content e-commerce platform, allowing users to complete hardware creation in just two steps: "describe → assemble."

· Future Plans:

· Short-term (1–2 years): Build the AI DIY creation platform prototype: achieve natural-language-to-3D-model generation and component recommendation; launch the VIBE Pack hardware component series covering electronics, mechanics, and structures; partner with MakerSpaces, universities, and K-12 institutions to capture the educational maker market.

· Medium-term (3–5 years): Build the Vibe Community where users can create, share, and replicate others' works; achieve a closed loop of online AI generation + offline smart manufacturing (3D printing, CNC, SMT fully automated production); launch AI smart toys and maker hardware kits for the family and parent-child creation market.

· Founding Team: Xie Jiahao (Lanson Xie), Founder & CEO. Master's from Harvard Graduate School of Design, teaching assistant at MIT Media Lab's maker space.

· Benchmark Brands: LEGO, Bambu Lab CyberBrick, Roblox.

· Business Model: SaaS + component e-commerce + platform creator incentive commissions.

· Funding Status: Seed round.

14 OU Tech

· One-liner: AI-powered scent customization and multi-sensory experience fusion.

· Company Overview: Founded in 2024, OU Tech is focused on AI scent customization and distribution through smart hardware. Its vision is to expand the boundaries of human sensory experience through frontier technology and culture, starting from the sense of smell.

· Future Plans: Reshape the cost structure and luxury narrative of the fragrance industry using OU's RoboShop scent customization technology and new offline distribution models, while revolutionizing people's olfactory sensory experience through new product forms featuring intelligent customization, contextualization, and audio-visual-olfactory integration.

· Founding Team:

· Founder Zhang Yufan: A second-year dropout from ShanghaiTech University, a serial entrepreneur with a hybrid background in technology, business, and philosophy.

· Founding team: Gen-Z dropouts with backgrounds in hardware/software, design, and perfumery.

· Partner & angel investor Yang Zhongqing: Previously achieved a successful entrepreneurial exit. As global executive manager at OneGarnetGroup, drove numerous top-tier client partnerships and locked in over RMB 500 million in profit.

· Benchmark Brand: PopMart.

· Business Model: Rapid expansion through a new store format featuring perfume customization experience robots, converting offline traffic at extremely low cost. Build the most extensive tagged database of fragrance users and the most precise user models, thereby incubating new olfactory experience products and ecosystem.

· Funding Status: Completed angel round in March 2025.

Recommended Reading

Where the Money Flows, Where People Rise and Fall