Building for exits in the AI era: a VC’s playbook

Scale Asia Ventures has recorded an impressive feat. Four years after the inception of its first fund, which raised US$30 million, the investment vehicle has already managed to see eight exits.

The San Francisco-based early-stage VC firm was founded in 2021, right before AI emerged as the defining technology for investors.

Notable exits for the fund include Promoted AI, which optimizes search results for ads and was acquired by Dropbox, and Argilla, an open-source data curation platform for large language models that was acquired by Hugging Face.

While the firm invests in US-based startups, its limited partners are primarily based in Asia, including conglomerates like South Korea’s Shinsegae Group and various family offices.

Scale Asia Ventures is now raising its second fund, says Wally Wang, founding managing partner of the firm. He declined to share a target size, but spoke with Tech in Asia about his company’s evolving investment focus and how it views exit strategies for AI startups today.

This interview has been edited for clarity and brevity.

How would you describe your investment ethos today, especially in the context of AI, where the market is moving so quickly?

We have been evolving our thesis. We started the firm back in 2021, so really at the beginning, it was about investing in B2B software. Then we gradually started to capture some opportunities in AI, especially AI B2B. But with the software development industry changing dramatically in the past three months or so, we have been shifting our investment thesis from just investing in the core software layer to two different areas.

One is going deeper into the deep-tech space, which includes investments in hardware, energy sources that support AI, and AI for sciences. The other direction we’re moving into is vertical applied AI applications, which cut across the legal space, customer support, healthcare, and financial services.

What are some of the startups in your portfolio that fall under this new investment ethos?

Under the deep-tech category, we’ve invested in d-Matrix, which produces a next-generation, low-latency, and energy-efficient inference chip. In terms of applications, we’ve invested in Cresta, an AI platform for call centers. We’ve also invested in Higgsfield, which is a fast-growing AI video creation platform.