Portfolio被投企业

Every company here had to answer the same question.这里的每一家,都要回答同一个问题。

Every company here is a direct investment, underwritten through one of two tracks and judged against the same test: can it become increasingly difficult for its market to replace? Each record carries its sector, its stage at entry and its current status — and where something checkable has happened, the date it happened and where you can verify it.

这里的每一家都是我们直接投资的,用两条评估路径中的一条来判断,检验的标准是同一个:它能不能变得越来越难被自己所在的市场替代?每条记录标明赛道、进入阶段与当前状态;凡是发生过可查证的事,都标出时间和查证途径。

25Companies家公司
12Software & applied AI软件与应用 AI
10Deep tech & sustainable industry深科技与可持续产业
3Life sciences生命科学
2Listed已上市
1Acquired被收购

Two tracks, one question underneath

两条评估路径,底下是同一个问题

The portfolio spans two distinct underwriting tracks. Software and applied AI compound through workflow ownership, habitual use and proprietary data. Deep tech and sustainable industry compound through new production capability, real-world validation and hard-to-replicate technical advantage. The evidence differs; the underlying question does not: can this company become increasingly difficult for its market to replace?

组合横跨两条不同的评估路径。软件与应用 AI 靠掌握关键环节、形成使用习惯、积累专有数据来复利;深科技与可持续产业靠新的生产能力、真实环境验证,以及难以复制的技术优势来复利。证据不同,底下的问题相同:这家公司能不能变得越来越难被它所在的市场替代?

12 companies12 家

Software and applied AI

软件与应用 AI

  1. Workflow
  2. habit
  3. data
  4. operating layer
  1. 工作流
  2. 习惯
  3. 数据
  4. 运营层
Where this path breaks这条路径断在哪里

This path compounds only when the software becomes where the work happens, not a tool beside it. It usually breaks earlier than people expect: the customer keeps their old system of record, so daily use never turns into habit and the data never accumulates.

这条路径要复利,前提是软件成为工作本身发生的地方,而不是旁边的一个工具。断掉的地方通常比想象中更早:客户留着原来的系统,于是日常使用没有变成习惯,数据也没有沉淀下来。

10 companies10 家

Deep tech & sustainable industry

深科技与可持续产业

  1. New capability
  2. real-world proof
  3. demand before scale
  4. hard to replicate
  1. 新能力
  2. 真实验证
  3. 规模化前已有需求
  4. 难以复制
Where this path breaks这条路径断在哪里

This path compounds only when a buyer commits before the capacity is built, because built capacity is what makes the advantage hard to copy. The step from real-world proof to committed demand is where most of these companies die.

这条路径要复利,前提是在产能建成之前就有人承诺买单——真正难被复制的,正是已经建起来的产能。多数这类企业死在从"验证过了"到"有人承诺买"的那一步,而不是死在技术上。

All companies

全部公司

Selected portfolio companies. Logos and company information are the property of their respective owners and are used with permission. Inclusion does not imply an endorsement of INP Capital, does not indicate investment performance, and does not imply operational control by INP Capital.

以上为部分被投企业。相关标识与企业信息归其权利人所有,已获授权使用。列示不代表该企业对 INP Capital 的背书,不代表投资业绩,也不代表 INP Capital 对其拥有经营控制权。

Portfolio news

被投企业近况

Recent developments at companies we have backed.

我们投过的公司最近在发生什么。

22 September 20262026年9月22日
Xpansiv
Announced a strategic partnership with Verdane, which is leading a capital raise to fund acquisitions and expansion into new products, asset classes and markets.与 Verdane 达成战略合作,由 Verdane 领投新一轮融资,用于并购以及新产品、新资产类别和新市场的扩张。xpansiv.com
18 September 20262026年9月18日
Power to Hydrogen
Began commercial green hydrogen production at the Port of Antwerp-Bruges, on the 500 kW electrolyser it delivered to the site in August.位于安特卫普—布鲁日港的 500 千瓦电解槽 8 月交付,9 月起开始商业化绿氢生产。power-h2.com
18 August 20262026年8月18日
GrubMarket
Acquired JR Holland, a Gateshead fresh produce and foodservice distributor serving North East England and Scotland — its first acquisition in the United Kingdom.收购 JR Holland——一家位于盖茨黑德、服务英格兰东北部与苏格兰的生鲜及餐饮食材分销商,是其在英国的首笔收购。grubmarket.com
23 June 20262026年6月23日
Wyvern
Selected by NASA for its commercial satellite data programme.入选 NASA 的商业卫星数据采购计划。wyvern.space