Key Takeaways

  • AI startups are hoovering seed capital, forcing pre-seed founders to clear seed-stage bars before they've built anything
  • The "Winning Pre-Seed Without a Product" panel at Disrupt 2026 puts three operators who've backed pre-revenue companies on stage to explain how conviction beats code
  • Venkatachalam, Agarwal, and Clements represent fundamentally different capital models — a new AI-focused fund, a 20-year enterprise veteran, and an inclusion-first accelerator
  • The session is a proxy war over whether storytelling still buys time in a market that worships velocity

The funding ladder has collapsed. AI companies raise seed rounds on Figma files and founder pedigrees while everyone else gets measured against traction metrics that didn't exist three years ago. Pre-seed founders aren't competing with each other anymore. They're competing with a category that moves faster, talks louder, and absorbs capital like a sponge.

TechCrunch Disrupt's Builders Stage understands this. The "Winning Pre-Seed Without a Product" panel isn't a feel-good workshop. It's a tactical briefing for founders watching the goalposts move in real time.

Sandhya Venkatachalam runs Axiom Partners, a $52 million fund built to connect founders with AI practitioners. She was first money into Groq. She backed GalileoAI, ForethoughtAI, FirefliesAI — all acquired or unicorns. Her thesis: the best pre-seed bets are founders who understand where AI actually improves workflows, not where it makes for good slides. She invests in translation skills. The ability to tell a practitioner what to build and why it matters.

Puneet Agarwal has seen this movie before. True Ventures started in 2005. Twelve funds. Five hundred portfolio companies. One thousand fifty founders. Sixty acquisitions. Seven IPOs. Agarwal has spent sixteen years watching enterprise infrastructure cycles turn. He knows the difference between a platform shift and a hype cycle. His pre-seed bar is operational: does this founder understand the procurement cycle, the security review, the integration pain that kills enterprise deals before they start?

Austin Clements runs Slauson & Co. Different scale. Different mission. Economic inclusion. Small business empowerment. He built an accelerator inside the firm because the founders he backs don't have Stanford networks or Y Combinator signals. They have revenue in markets VCs ignore. Clements' pre-seed conviction comes from cash flow, not narrative. He funds the unsexy companies that AI startups will eventually automate.

Three capital allocators. Three different conviction frameworks. That's the point.

The panel title promises a playbook. "Winning Pre-Seed Without a Product" sounds like a hack. It's not. The real lesson is that product was never the prerequisite investors claimed it was. Traction was. Revenue was. Insight was. AI just made product cheaper to fake, so the real signals — obsession with a specific problem, access to a specific customer, clarity on why now — got more expensive to verify.

Venkatachalam will argue that AI-native founders win by demonstrating they can direct technical capacity toward real workflows. Agarwal will argue that enterprise founders win by proving they've mapped the buying committee. Clements will argue that overlooked founders win by showing recurring revenue from customers nobody else respects.

None of them will say "just have a great story." They'll say: have a story that survives cross-examination.

The skepticism is warranted. Most pre-seed pitches are delusion wrapped in TAM math. The panel exists because the noise floor has risen. Founders need to know which frequencies actually transmit.

Disrupt 2026 runs October 13-15 at Moscone West. The Builders Stage sessions are where operators talk to operators. No keynote fluff. This panel is the one that matters for anyone raising before product-market fit in a market that pretends product-market fit arrives at incorporation.

Get the ticket. Sit in the room. Listen for the differences between the three speakers — not the overlaps. The overlaps are marketing. The differences are where the actual criteria live.