Seed-focused venture capital firm Kae Capital, known for backing companies like Porter, Zetwerk and 1MG, is doubling down on themes such as AI-led automation and deeptech sectors like manufacturing, energy transition, and defence, even as it warns of excesses building up in parts of the market. In an interview with Ayanti Bera, General Partner Abhishek Srivastava speaks about the surge in AI funding, how seed-stage economics are changing, and how current geopolitical tensions can impact venture funding. Excerpts:
What is Kae Capital’s investment strategy today?
We’ve always been a very conviction-led, first-check investor, where we focus primarily on pre-seed and seed. Today, we broadly operate across two anchor themes. One is AI-driven intelligent automation, including both B2B and consumer AI. The second is what we call “Resilient India”, which includes manufacturing, deep tech, energy transition, defence, and supply chain resilience.
You’ve backed several consumer brands in the past. Are you still bullish on the consumer space?
Yes, but selectively. Consumer has been very kind to us, companies like Traya, Foxtale, and earlier investments like 1MG and Porter have created strong outcomes. But going forward, we’ll be selective in consumer and fintech.
There’s a lot of buzz around AI and deep tech. Are you seeing overcrowding at the seed stage?
Absolutely, an amazing level of overcrowding. It’s a great time to be an entrepreneur, especially in AI and deep tech. But from an investor perspective, it demands caution. We prefer looking at “picks and shovels” opportunities, that is, companies that enable others in a booming sector. For example, instead of backing every semiconductor startup, we’d rather back tools that help multiple semiconductor companies scale faster.
How has this hype impacted valuations at the seed stage?
Seed cheque sizes have definitely gone up. AI companies today are raising $3-5 million at seed, which earlier would have been closer to a pre-Series A or Series A round. Moreover, there’s ample capital supply, which creates some artificiality in valuations. That said, traditional sectors still see more disciplined seed rounds in the $1-2 million range.
Within AI, which segments are attracting the most capital?
Vibe coding tools have seen significant attention, though I believe this space will consolidate into one or two dominant players. Other emerging areas include Voice AI, especially for India’s regional languages, AI companions, although it’s still very early, and AI-powered B2B services. Among these, AI services, in particular, could become a major opportunity
Why do you think AI services could be the next big trend?
Historically, Silicon Valley investors were skeptical of services models, but that thinking is evolving. India has a natural advantage here, that is, talent availability, cost efficiency, and the ability to scale service-led models. Over the next 6–12 months, AI services could emerge as a major funding theme.
How are macro factors like geopolitical tensions affecting venture capital?
In the short term, there will always be volatility. But VC operates on a 10-year cycle, so structural changes matter more than temporary ones. What we will see, however, is that not all funds will survive. There will likely be a few “zombie VCs” that fail to deliver returns. Ultimately, capital will flow to managers who can demonstrate consistent outcomes.
What about exits? Are you concerned about valuations going forward?
Near-term IPO plans may get delayed due to market volatility, but fundamentally, strong companies will still create value. For early-stage investors like us, we usually have enough cushion. We also rely on secondaries in later rounds to return capital. India’s capital markets remain a strong exit avenue, though M&A activity still needs to mature further.
Has the seed-to-Series A journey become tougher for startups?
Yes, timelines have stretched compared to the 2021–22 period. Investors are now far more focused on discipline, clear paths to profitability and sustainable growth. In AI, it’s even more uncertain because the industry is still evolving. We’re yet to see what a “true” Series A or Series B looks like in AI and here the benchmarks are still being defined.