# PKSHA Capital

PKSHA Capital is a Private Equity based in Tokyo, Japan.

## Overview

- **Organization type:** Private Equity
- **Headquarters:** Tokyo, Japan
- **Region:** Asia
- **Address:** Hongo Segawa Building 5F, 2-35-10 Hongo, Bunkyo-ku, Tokyo, Japan
- **Assets under management:** Undisclosed
- **Website:** pkshacapital.com
- **LinkedIn:** https://linkedin.com/company/pksha-algorithm-fund

## Regulatory record

- **Reports private funds:** No

## About

PKSHA Capital is a subsidiary of PKSHA Technology. It manages PKSHA SPARX Algorithm Investment, a fund focused on Japanese and foreign start-ups with AI and robotics technology. The fund is jointly managed by PKSHA Technology Capital and SPARX AI & Technology Investment, a SPARX Group subsidiary.

## Sectors

- AI/ML
- Enterprise Software
- Robotics & Automation
- Mobility & Transportation
- Digital Health
- FinTech
- Energy Transition & Renewables
- PropTech
- AgriTech & FoodTech
- Media & Entertainment
- Cybersecurity
- Education
- Industrial Tech
- Healthcare Services

## Offices

- Seoul, South Korea
- Shanghai, China

## People

- Hideyuki Ebihara — Investment Team
- Shunsuke Itadani — Investment Team
- Tengqi Hu — Investment Team
- Maowen Cheng — Investment Team
- Sanghyun Han — Investment Team
- Kangsoo Kim — Director, Technology Team / CEO, MK Capital

## Questions

### Who runs investment decisions at PKSHA Capital?

The investment team is led by Hideyuki Ebihara, who spent roughly 12 years at CyberAgent Ventures (now CyberAgent Capital) and led the firm's Korea office launch from Seoul. He works alongside senior investment professionals Shunsuke Itadani (formerly China head at Dream Incubator), Tengqi Hu, Maowen Cheng, and Sanghyun Han, who collectively cover Japan, China, Korea, and Southeast Asia with local operating experience at firms such as Mizuho Securities and Nomura Research Institute.

### How does PKSHA Capital source proprietary deal flow?

Sourcing draws on a structural advantage: the firm's strategic partnership with the University of Tokyo's Matsuo Laboratory provides early visibility into research-originated AI startups. The multi-office footprint in Tokyo, Seoul, and Shanghai allows on-the-ground presence in three major Asian technology hubs. Additionally, team members' prior roles at CyberAgent Capital, Dream Incubator, and local consulting firms create networks that span the Japanese, Korean, and Chinese startup ecosystems.

### Is PKSHA Capital structured as a corporate venture arm or an independent fund?

It operates as the external early-stage investment entity of PKSHA Technology, a Tokyo Stock Exchange-listed AI software company, but with operational independence. It maintains its own investment team, brand (PKSHA Algorithm Fund), and portfolio, while benefiting from the technical resources of its parent and the academic pipeline of the Matsuo Laboratory.

### Which sectors does PKSHA Capital explicitly avoid?

The firm's portfolio concentrates on AI and software applied to industrial transformation. Sectors notably absent from its published portfolio include deep biotech therapeutics (beyond AI-enabled drug-trial support from Hope AI), pure-play hardware without a software layer, and traditional heavy-industry plays without a technology-enabled automation thesis.

### How is PKSHA Capital related to the University of Tokyo's Matsuo Laboratory?

PKSHA Capital lists the Matsuo Laboratory as a strategic partner, providing a direct channel to academic research in deep learning and social implementation of AI. Professor Yutaka Matsuo's lab is a wellspring of AI talent in Japan, and the relationship gives the firm access to spinout opportunities and technical due-diligence capabilities that generalist VCs lack.

## Related profiles

- [Pkrids Wealth Management](https://altss.com/profile/pkrids-wealth-management)
- [Pku Capital](https://altss.com/profile/pku-capital)

---

Last updated: 2026-06-03T20:00:00.000Z

Canonical page: https://altss.com/profile/pksha-capital

Maintained by Altss — https://altss.com — methodology: https://altss.com/methodology