In this episode of The Sunny Ray Show, host Sunny Ray talks with Vik Ghai of G2C Ventures, an early stage venture firm based in Palo Alto whose name stands for grit to capital. Ghai explains that after multiple startup exits in physical security and manufacturing, he chose to support founders rather than start another company, and now backs a portfolio of around 40 companies. He describes G2C's investment thesis as a system of outcome: AI driven software that rebuilds a specific vertical workflow, such as construction or physical security, and charges customers for results rather than licenses or seats. Ghai argues that large language models have become commoditized, comparing a model to flour in a cake recipe, meaning the real value lies in the full application layer built around it, including a human in the loop for reliability. He also discusses what it takes to earn a first meeting, why referrals produce his best deals, and how G2C evaluates founder grit within the first few minutes of a conversation. The conversation closes with his priorities for the next quarter of fundraising and deployment.
G2C Ventures' Vik Ghai on why AI models are just flour and outcome-driven vertical software is the real cake.
Tell us about G2C Ventures. What's unique about how you see the world?
G2C stands for grit to capital. We're an early stage venture firm based in Palo Alto. We see too many ideas, so we look for founders with grit, not just ideas. Grit means founders willing to slog it out, survive near death experiences, and keep going, and that takes more than capital. Founders bring the grit, we bring the capital, and that's a great combination.
You built companies in physical security and manufacturing. What did those markets teach you that software people usually miss?
Software people often think the job is done once you ship, but in physical security and manufacturing the real work starts when people begin using it and building custom workflows. You are integrating software with devices, so that's where the rubber meets the road. Deployment is harder than people expect, and those industries taught me the resilience software needs for constant real world change.
What was the hardest operational problem you personally had to solve as a founder?
The hardest part wasn't building the software, it was getting five departments to agree it was the right solution. Even if security loved it, I still had to bring HR and procurement along, and people operate very differently across continents. Getting corporate America, Europe and Asia to agree on one solution was the real challenge, especially when trying to scale beyond the first few sites.
After your exits you hit a fork: build another company or build founders. Can you walk us through that decision?
After a couple of exits I could have started another company, but I decided to use my experience to support 40 founders instead, which is exactly what G2C Ventures does. We work like Silicon Valley culture demands, essentially around the clock, and being a sounding board for 40 founders having the same conversations I once had with myself felt like a bigger cause.
You said you look for companies that absorb the services layer. Why is that the hard part?
Services are the messy part most software companies avoid, but AI is probabilistic while corporates want deterministic, near correct answers. We invest in end to end automation with a human in the loop by design, not as an afterthought, like a construction RFP tool that also provides structural engineer review so the output is guaranteed to be structurally sound.
You said the model is flour. When did that land for you and what did it change?
It became clear in 2025 as model companies started offering similar base capabilities through common APIs, so the model alone doesn't solve the problem, it's just an ingredient. Referencing a five layer AI cake, people want the whole cake, the application, control plane and infrastructure together, priced as one outcome. That's why our thesis is rebuilding the workflow, the cake, not the flour.
What does it feel like when you and a founder click, can you tell inside of ten minutes?
We wouldn't write a check in ten minutes, but you can tell how open a founder is right away, through their personal story and the grit they show. I especially click with founders who talk about a real customer's problem rather than just their own solution, because relating their product to a real company's need gives me strong validation we'll click long term.
What are the next 90 days about for G2C, and what has to be true by the end of them?
September through November is where most deal activity happens, so we're meeting as many founders as possible to close a couple more deals before year end. We have capital allocated and are also actively fundraising for our current and next fund, since venture is a business where you deploy and raise capital at the same time continuously.
Building something daring? Sunny talks to founders like this every day. Fifteen minutes to see if your story belongs on the stage.
Claim your pre-interview