Jennifer Savage, Partner Illuminate Ventures
Jennifer Savage is a Partner at Illuminate Ventures, a B2B Seed-stage venture capital firm behind early successes like Xactly, Coupang, BrightEdge, and Contentstack. She leads new investments, advises portfolio companies on strategy and operations, and serves on the boards of Brevity, TAZI AI, Pyze, and mpathic. A seasoned B2B software executive with over 25 years of experience, Jennifer has built new categories and led successful product and go-to-market strategies across multiple high-growth companies. Prior to joining Illuminate, she held executive roles at PlaceWare (acquired by Microsoft), DocuSign (IPO), Smartsheet (IPO), and Flowroute (acquired by West Corporation). Jennifer earned her MBA from UCLA’s Anderson School of Management and holds a bachelor’s degree in Computer Science from the University of Oregon.
Xuhui Shao, Managing Partner Foothill Ventures
Dr. Xuhui Shao is the Managing Partner of Foothill Ventures, a technology-focused venture fund investing in early stage startups in the US. His investment areas are AI applications, enterprise software and full-stack sensor/data/algo companies. Before that Xuhui has been Yahoo’s VP-II Engineering in charge of Yahoo’s global display ads platforms; Founding technologist of Ad Tech pioneer company Turn ($310M exit) and AI risk management company ID Analytics ($170M exit). Xuhui earned his BS and MS degrees from Tsinghua University, and his Ph.D. in EECS from University of Minnesota. Xuhui is an early investor of Plus.AI, Weride.AI and Certik – all subsequent unicorn compa
Investment Focus
Jennifer: I’m Jennifer Savage, one of the partners at Illuminate Ventures, and we’re seed stage enterprise software investors and investing out of our third fund. And we have a preference for what we call further along seed stage companies. We think it’s an enduring trend that you can get companies off the ground with less capital, unless, of course, you’re building an LLM. And we like to invest in those companies, that relative to the capital they’ve raised, they’ve gotten further along in terms of revenue, team, product or other advantages.
Xuhui: I’m Shihui Xiao, one of the Managing Partners at Foothill Ventures. We are an early stage deep tech venture fund based in Silicon Valley. So our funds tend to pick early stage, deeply technical teams in an area that we believe have disruptive technology and can have a huge impact on a giant industry and sometimes multiple industries. That leads to cross-disciplinary innovative opportunities. But for me personally, I tend to spend time mostly in the AI enterprise software area, and especially the last few years where AI has been accelerating its pace in innovation as well as in the startup world. So nowadays, I spend most of the time in AI, whether it’s AI infrastructure or vertical applications such as robotics and legal tech and a few other examples.
Defensible AI Moats equal Viable Businesses
Jennifer: I like to think of a moat as something about the solution that gets better over time in a way that’s hard for other competitors to replicate. And better over time, meaning delivering more and more value to the customer. Yeah, it’s hard. you know and it’s not good enough just to win customers, you’ve got to keep them. And so if they’re getting happier and happier, it’s very hard for them to leave.
Xuhui: So I think a moat is something we think about as VC investors all the time. And at the end of the day, a startup is only valuable if they can make something impossible possible, which means either they have to invent new innovation, new science, new algorithm, or they have to make existing technology stretch far enough to do something seemingly magical, impossible! And so that’s especially interesting in the AI world where things are suddenly changing, two and a half years ago when ChatGPT launched! And then we’ve been thinking about this thing ever since.
Impact of DeepSeek
Xuhui: So when ChatGPT first launched, we proposed a framework that’s not unique to us. All those similar things I look at the foundational AI LLM world.by breaking it down into algorithms, compute, and data – the three dimensions. And I propose that small startups can only focus on the third dimension, which is data, because they cannot really own or dominate the first two. Now, there’s always exceptions! DeepSeek is one of the major exceptions! They actually made a breakthrough as a small company on algorithms and compute. Now, its breakthrough shifted the focus from pre-training to post-training, yet it did not change our view of the algorithm: compute, data, framework. We still believe that even though the focus has shifted from pre-training to post-training, it still fits the original hypothesis that most of the startups focus on data as the competitive moat.
OpenAI’s Apparent Dominance
Jennifer: Well, it’s interesting you frame it that way, because I think that most of us were surprised at how much traction and how well-performing the other LLMs have done as compared to OpenAI, because when it first came out, it was this perception that they’ve just run away with it. And I think of it more as what’s driving OpenAI is the first mover advantage and the end user adoption and awareness that is helping to fuel the overall business. But I think that what we’re seeing more and more in platforms adopting foundation models is, they’re making sure that they’re foundation model agnostic.
Xuhui: OpenAI first succeeded as a horizontal play that really shocked everyone. And then technologists really prefer horizontal play as it is often technology first. and then takes less market risk and easier to pivot. So OpenAI were able to pivot from a video generation or image generation to text generation to become a platform, to enabling agentic AI and so on and so forth.
Now, on the other hand, you see that not only has this spurred other people to compete in this area, there’s also an internal diffusion rule in that the original founding team, most of them have left to start other companies, right? So Anthropic, Thinking Machines Lab and others are all funded by former OpenAI founders or early employees.
Jennifer: Yeah, I agree with that. I think that the interesting thing is that the advantage often for the vertical plays is inherently in the AI itself. And in terms of specialized knowledge, specialized data, all the things that you had mentioned. And a lot of times in the horizontal world, when you’ve got the advantage it comes at least partially in the go to market, which may or may not have an AI component. And it is a distribution advantage which often plays a bigger role in horizontal plays or it can? And so it ends up being, it might be a better opportunity in terms of larger TAM, but it might not be the AI that’s actually driving the winning formula for that type of horizontal play.
Silicon Valley Advantage
Xuhui: So this is the magic of Silicon Valley in that people move around and then, it creates competing solutions and in the end that really pushes the acceleration of any innovation. Now, coming back as an investor, I often prefer vertical AI because for startups, it’s more focused. It’s easy to demonstrate the value and easy to adopt. And then for the last 12 months, there has emerged a more important factor. The development costs were dropping. We’re seeing one person team, two person team, are able to create end-to-end AI applications that offer a lot of value to their customers. So the value of apps are increasing while the development costs are decreasing or dropping significantly. So logically, this trend or these two trends in combination start to favor more customized vertical solutions as opposed to one size fit all. Because you can afford it and it can show tremendous value! So for these reasons, we’re still very bullish and more bullish about vertical applications. We think this would be a huge replacement cycle of the SaaS technology wave where all the SaaS companies become AI companies or all the SaaS user cases become AI user cases rather.
I feel I’m also transplanted to Silicon Valley and slightly earlier than Pemo, in 2010. But I think everybody feels they are transplanted here. There are very few natives here, right? That’s the magic of Silicon Valley in that it’s enduring ability to absorb changes, to sustain changes, and to promote changes, right? So to me, it’s hard to say whether the recent changes are good or bad or even harder to weigh the so-called good against the so-called bad, right? So I think the resilience of Silicon Valley has been demonstrated over and over again from every wave of innovation from web 1.0 to 2.0, to mobile web, to AI 1.0 2.0, to now generative AI, and then beyond. So I think it’s really demonstrated as such a deep, rich ecosystem of research universities, big tech companies, vibrant startup ecosystems, and the VCs all working together! So I’m really excited! There’s no coincidence that the AI revolution started here and will continue to balloon in Silicon Valley. Because all the talent, all the Capital, and all the right kind of setup is densely packed in, in this small piece of land.
Jennifer: Well, I come from a different perspective in the sense that Illuminate is based in the Bay Area, although I personally live in the Seattle area. And I guess what I would observe, with one foot in multiple camps, is that with the rise of AI, it is the perception of Silicon Valley that Venture Capital has risen positively because of the level of innovation. I think it’s reached public knowledge in a way that technology coming out of Venture Capital in general just never has before. And it’s received positively. I mean, people love it! And I know there’s questions and all of that? So I feel like this is a situation where all boats are rising on the wave of AI, not just Silicon Valley, but including Silicon Valley.
