What does it take to leave a senior engineering role at Google and start an AI consulting company from scratch? In an in-depth interview with Tech Startup Network, Krazimo founder Akhil Verghese tells the full story.
Verghese’s journey began at BITS Pilani in India, where he studied physics and civil engineering before pivoting to software. After starting at Fiberlink (later acquired by IBM), he spent years as a machine learning consultant and served as the founding Head of AI at Butter.ai, a startup backed by General Catalyst. In 2019, he joined Google, where he spent six years — ultimately leading reporting projects for Gemini within Google Workspace and advising teams on optimizing LLMs for reliability.
That advisory work is what sparked Krazimo. Verghese saw firsthand how even sophisticated companies struggled to deploy AI reliably in high-stakes environments. The gap between a compelling demo and a production-ready system was vast, and most organizations lacked the engineering discipline to bridge it.
The interview covers Krazimo’s philosophy of enterprise-grade AI: systems that are creative and intelligent yet remain predictable, testable, and auditable. Verghese explains the company’s signature phased launch strategy — shadow launches, human-in-the-loop validation, and only then full automation — and discusses why engineering rigor matters more than ever in the age of generative AI.
Originally published on Tech Startup Network. Krazimo specializes in reliable, enterprise-grade generative AI solutions built by former Google engineers.
Read more on the Tech Startup Network.