Shomik Ghosh is a Partner at Sierra Ventures. He was previously a Partner at Boldstart Ventures, where he focused on investing at inception in technical founders building products to solve enterprise pain points, such as Cloudquery, Kiln AI, and Noded AI. Prior to that, he was a growth-stage investor at Top Tier Capital Partners, investing from Series B through pre-IPO in companies such as CircleCI, Anaplan, and Shape Security.
He previously has experience in tech M&A, fixed income trading, venture debt, and at a failed startup.
He's an avid fan of hiking and enjoys spending time with his young son when he has free time. He is known as the cautionary tale of playing pickleball for fun, not stretching, and ending up with a torn Achilles.
What parts of AI feel underexplored right now?
What does a “frontier AI founder” look like today?
Frontier founders usually start with a vision that feels unrealistic at first, then clearly map the technical steps to get there. They’re pushing the edge of what exists today—whether in models, chips, or robotics—to accelerate a future that otherwise would take much longer.
What are founders getting wrong about building in the AI era?
AI will reinvent a lot, but many fundamentals still matter. Great sales, strong hiring, and customer value still win—the difference now is teams that combine human judgment with AI will outperform those relying on either alone.
How should founders and investors think about TAM (total addressable market)?