A deep dive with three leading AI investors who collectively manage billions in venture capital and have backed some of the most innovative companies in artificial intelligence.
The Expert Panel
Karen Paige, General Partner at B Capital Karen brings deep operational and investment expertise to B Capital, where she focuses on enterprise software and AI investments. B Capital manages over $6B in assets and has backed category-defining companies across enterprise software, fintech, and healthcare.
Rudina Seseri, Managing Partner at Glasswing Ventures As founder and Managing Partner of Glasswing Ventures, Rudina leads investments in early-stage AI and Frontier Tech companies, with a particular focus on Enterprise and cybersecurity markets. She brings over two decades of experience in technology venture capital and has been recognized as one of the Top 100 Most Influential Women in Business.
Joselyn Goldfein, Managing Director at Zeta Venture Partners Joselyn is a technologist-turned-investor who leads AI and data infrastructure investments at Zeta Venture Partners, focusing on inception through seed stage startups. Previously a Director of Engineering at Facebook and VP of Engineering at VMware, she brings deep technical expertise to evaluating early-stage AI companies.
The State of AI Adoption: Mind Share vs. Wallet Share
Let’s be real – there’s a massive gap between AI hype and actual enterprise adoption. While 50% of enterprises have dabbled in AI projects, most are still playing around the edges rather than transforming their core business. Here’s what’s really happening:
The Great AI Experiment
Every boardroom is talking about AI. Public company earnings calls are peppered with AI mentions. But when you look under the hood, what you find is mostly experiments and POCs. One top-five global bank’s chief data officer admitted they had 150 GenAI projects in the lab – but zero in production. That tells you everything you need to know about where we are in the adoption curve.
Why Enterprises Are Hesitating
Despite the FOMO driving interest, enterprises are gun-shy about major AI investments for several key reasons:
- Unknown risks around data and compliance
- Complexity of implementation
- High costs of data infrastructure
- Talent gaps in AI/ML
What This Means for AI Founders
The Infrastructure Trap
2023 was the year of demos. 2024 needs to be the year of production deployment. But here’s the challenge – we’re seeing too many founders fall into what I call the “infrastructure trap.” They’re building glorified wrappers around OpenAI’s APIs and calling themselves AI companies. That’s not going to cut it.
Where the Real Opportunity Lives
The biggest opportunities right now are in what I call “AI applications as a service” (AIaaS). These are purpose-built applications that:
- Solve specific departmental problems
- Target users 1-2 levels below the C-suite
- Drive measurable ROI
- Deploy through bottom-up adoption
The Three Keys to Building a Defensible AI Company
1. The Data Moat
Your data strategy needs to go beyond just having a lot of data. The most successful AI companies we’re seeing are building moats through:
- Unique data sources (e.g., wet lab data in life sciences)
- Superior data curation and annotation
- Narrow vertical focus leading to higher quality data
- Network effects that improve with scale
2. The Distribution Play
Don’t try to go head-to-head with incumbents just because you have AI. They have distribution and mountains of data. Instead:
- Find greenfield opportunities where AI enables previously impossible solutions
- Build new workflows rather than just AI-enabling existing ones
- Target departments or functions where bottom-up adoption can flourish
3. The ROI Story
The winners in AI will be able to articulate crystal clear ROI. Take FeatureByte – they can tell prospects: “This process normally takes 6 months and $2M per dataset. We can do it in X time for Y cost.” That’s the level of clarity you need.
Go-to-Market Strategies That Actually Work
The PLG Motion in AI
Product-led growth isn’t dead in AI – it’s evolving. The most successful motions we’re seeing:
- Target individual practitioners first (e.g., developers, data scientists)
- Focus on immediate productivity gains
- Make onboarding frictionless
- Build viral loops into core product use
The Vertical Approach
Specialization is winning, especially above the infrastructure layer. Companies like Laiva have become the go-to platform in life sciences by:
- Deeply understanding vertical-specific workflows
- Building purpose-built AI for industry use cases
- Creating two-sided marketplace effects
- Driving high annual retention through workflow lock-in
What VCs Are Really Looking For
The MVP Question
The definition of MVP in AI is situational. You need to understand:
- Required accuracy levels for your use case
- Acceptable error rates
- Time to value
- Implementation complexity
The Platform Vision
While you need to start focused, VCs want to see:
- Clear expansion paths
- Network effect potential
- Data accumulation advantages
- Logical adjacent markets
The New Metrics That Matter
Traditional SaaS metrics matter, but AI companies need to show:
- Data advantage metrics
- Unique data points
- Annotation quality
- Model improvement rates
- Usage depth metrics
- Time saved
- Error reduction
- Process automation rates
- Economic metrics
- Cost per inference
- Margin expansion with scale
- Data acquisition costs
Key Takeaways for Founders
- Don’t build infrastructure unless you have a truly novel approach. The market is saturated.
- Focus on specific problems rather than general-purpose AI. The more focused, the better.
- Build for production, not demos. 2024 is about getting real deployments.
- Think about governance early. Enterprises need it to scale AI.
- Create clear ROI stories backed by data. The days of selling AI on potential are ending.
The next great AI companies won’t win just because they use AI. They’ll win because they solve real problems in ways that weren’t possible before, with clear ROI and defendable advantages. Focus there, and you’ll have a much better shot at both customer adoption and VC dollars.
