Forget the Old Playbook: How Data/AI Founders and Investors Win Now
Best way to figure out what investors are thinking? Ask - over a friendly breakfast.
So we did just that. Over breakfast in Tel Aviv for early stage founders, we brought together three investors for an insider’s look into what the playbook looks like for investing in data and AI companies today:
- Erica Brescia, Managing Director at Redpoint Ventures
- Moran Levinovitz, Group Head of Ventures at HSBC
- Guy Fighel, Venture Partner & Head of Data Program at Hetz Ventures
Each brings a different vantage point — from seed-stage technical investing to corporate venture to later-stage infrastructure bets — offering a unique look into how leading investors are thinking about this fast-moving space.
The conversation ranged from how they identify emerging trends, to what founders often get wrong about competition and defensibility, to how the evolving AI landscape is reshaping what it takes to build lasting companies.
Here are the key takeaways for founders and builders operating at the frontlines.
How VCs Actually Spot Real Trends and Avoid Hype
Despite their different backgrounds, all three investors emphasized the importance of staying close to real technology and users — not just market narratives.
Guy Fighel described a highly technical, hands-on approach. As a former operator, he still codes weekly, tests products directly, and digs into open-source repositories. His goal is to understand firsthand what works, what’s half-baked, and where real technical gaps exist. This direct engagement allows him to separate surface-level hype from true engineering opportunities long before they become obvious to the broader market.
"I code at least once a week. I dig into the repos. I want to know if the library even works, if it’s half-baked or actually usable.” - Guy
Erica Brescia, who spent 20 years scaling open source and infrastructure companies before moving into venture, takes a slightly different but complementary angle. At Redpoint, trendspotting often comes from staying close to their portfolio companies — the builders creating next-generation systems. Tracking what cutting-edge technical founders are adopting in their stacks often reveals the earliest signals of new shifts.
Meanwhile, Moran Levinovitz offered the enterprise perspective: with HSBC’s global technology organization — tens of thousands of technology professionals — they can internally validate startups not just for functionality, but also for enterprise readiness. At their scale, it's not enough for a tool to work; it must withstand the compliance, resilience, and security demands of a highly regulated environment.
Bottom line: In today's market, investors aren’t just relying on buzz. They expect founders to have technical depth, user insights, and a grounded understanding of what it really takes to build enduring systems.
Where the Smart Money Is Going in AI and Data
When it comes to identifying the next big opportunities, all three speakers pointed to nuanced but powerful shifts — not surface-level hype cycles.
Guy focused on a strategy that many founders underestimate: building with the hyperscalers, not against them. Rather than trying to outcompete giants like Databricks, Snowflake, or AWS directly, smart startups can thrive by building complementary solutions that fill the gaps left by their roadmaps. Understanding what a major platform is unlikely to prioritize — and aligning with their ecosystem — can open powerful go-to-market channels and reduce existential risk.
Erica pointed to an important architectural transition: the move away from expensive, tightly structured systems like Splunk toward lighter, S3-based storage backends with smarter application layers. In addition, she highlighted the emerging need for new infrastructure to support AI agents — systems that will increasingly need live, automated access to data in order to take actions. While the future AI stack remains unclear, the winners will likely enable or redefine how data is processed, accessed, and activated at scale.
"We’re still figuring out how agents will actually get access to data and take action. There’s white space there for new kinds of tooling." - Erica
From the enterprise side, Moran outlined a dual horizon. In the next two to three years, success will come from bridging AI into enterprise environments with the right guardrails: compliance, cyber risk management, model governance, and regulatory alignment. Longer-term, Moran emphasized that global forces like compute costs, and energy access will reshape which players can afford to lead in AI.
Bottom line: The biggest opportunities aren’t always obvious. They're often found in the second-order consequences of platform shifts, enterprise needs, and macroeconomic forces.
What Gets Startups Killed Before They Scale
Just as important as spotting what's exciting is understanding where founders often go wrong.
Across the board, the investors agreed that incrementalism is not enough. Erica pointed out that too many startups are focused on making small improvements to how work is currently done, rather than reimagining how work itself will evolve with AI. Building a 20% better system won't create lasting advantage; founders need to think on a first-principles level about how foundational models, automation, and new interfaces will transform industries.
Moran warned against building products that look like features rather than platforms. Enterprises, he noted, can only engage with a limited number of vendors, and are seeking broad, integrated solutions that can consolidate multiple workflows. A startup pitching a narrow tool risks being outcompeted by incumbents who can easily add similar capabilities.
"Enterprise buyers can only deal with so many vendors. If your product looks like a feature rather than a platform — it’s a tough spot to be in." - Moran
Guy emphasized two critical gaps he sees repeatedly among founders. First, many underestimate the competitive landscape, failing to deeply research or differentiate themselves against the true field of competitors — not just the two or three they are aware of, but often dozens. Second, founders often ignore the existential threat posed by hyperscalers. If AWS or Snowflake can recreate your product with a handful of engineers in a single quarter, your long-term prospects are fragile unless you build in defensibility from day one.
Bottom line: Investors are looking for companies that can define new categories, not just improve existing processes. Defensibility — against both startups and giants — must be built into the company’s DNA early.
How to Build VC Relationships Before You Even Fundraise
Another major theme was the importance of building strong, strategic relationships with investors early — but doing so thoughtfully.
Moran noted that for corporate VCs and strategic investors, timing matters. Engaging too early, before a startup is ready to handle enterprise procurement and compliance, can waste valuable founder time and resources. Founders should be realistic about when they’re enterprise-ready — and manage the engagement accordingly.
Erica challenged the common wisdom that founders should only approach investors when actively raising. She shared that Redpoint often builds relationships with founders one to two years before investing. These early conversations allow the fund to track a company's progress, deepen conviction, and sometimes move quickly when an opportunity arises. Founders who proactively build relationships — without hard pitching too early — often position themselves for a stronger raise later.
"It's a 10-year journey when we invest. Meeting well ahead of when you are actually going to raise can put you in a stronger position later. The VC-founder relationship is so important." - Erica
Guy offered a practical tip for gaining early credibility: build something, no matter how lightweight, and put it in the hands of potential users. Even unpaid pilots with real enterprise logos can shift the entire tone of a conversation with investors. Demonstrating traction — not just potential — signals seriousness and de-risks the opportunity early.
Bottom line: Relationship-building with investors should start early, be strategic, and be supported by tangible evidence of progress, not just ideas.
AI Lowers the Bar for Building - But Raises the Bar for Winning
While AI tools have lowered the barrier to product development, the panelists agreed that execution beyond the codebase matters more than ever.
"Technology itself is not the business. It’s how people use it to create new opportunities." - Guy
Strong technical teams are now table stakes. Winning companies will be defined by their ability to build go-to-market engines, establish distribution channels, and create ecosystems where switching costs are high. As Erica Brescia put it, success will come from embedding into critical workflows and owning data or operational layers — not just by building technically superior products.
Moreover, as AI commoditizes many technical capabilities, other factors — like branding, partnerships, regulatory readiness, and speed of execution — will increasingly determine which companies break out.
Bottom line: The AI revolution changes what’s possible — but not what it takes to win. Great products are essential, but great companies are built through relentless execution across technology, distribution, and strategy.
Founders building in the AI and data space today are operating at one of the most dynamic moments in technology history. But winning won't come from chasing hype cycles. It will come from deep technical understanding, strategic clarity, defensibility against incumbents, and operational excellence.
Thanks to our partners at Redpoint Ventures and HSBC Innovation Banking for a fantastic breakfast together in Tel Aviv.
