.svg)
.svg)
%203.svg)
.png)
.svg)
Lesson from my journey
.svg)
Soumya Yadav
The best companies are not built on ideas; they are built on a founder's refusal to accept that a problem has to stay broken.
Soumya spends his time looking for that refusal, for founders whose reading of a market is earned through years of living inside the problem, rather than inferred from a deck. As a technology investor in AI/SaaS, enterprise software and deep tech, he is drawn to working at the edge of what technology can reliably do today, founders those who design around constraints instead of assuming them away. The ones who understand where product breaks and tend to build processes that don't.
He believes venture is an apprenticeship on both sides, that the best investors earn their seat not in the first meeting but in the hundred that follow, and that the founders worth backing want a partner who will show up for that work. Unglamorous bets. Compounding ones.
Soumya’s experience spans AI research, operating roles in large-scale technology business, and technology venture investing.
His last decade in technology has moved between research (AI) and operations. At IIT Kharagpur, he worked on recommendation systems, computer vision (monocular object detection), and IoT+ML systems, with publication in peer-reviewed journal. This technical grounding extends into his current sector focus across AI systems and emerging quantum technologies. He then moved as an operator at Flipkart, he ran a $180M category through one of its fastest growth phases. The remit was end-to-end; identifying where growth pockets or whitespaces were available, translating it into the demand plan, driving execution against it, and owning the financial model and category margins underneath. Growth was driven through marketing, modelled through the P&L, and audited against margin at every step. These experiences clarified a distinction that now shapes how he invests at Kalaari: technically impressive and commercially durable are different things.
Outside of work, he enjoys scuba diving, paragliding, and surfing, and most Sundays, you will find him cafe-hopping in Indiranagar or Koramangala.
Favourite quote: “You are remembered for the rules you break.” — Phil Knight
In most technology cycles, the underlying capability layer becomes commoditized over time, and value shifts toward products that integrate that capability into workflows.
In AI and enterprise software, this translates into a shift from model performance to workflow ownership, system integrations, and reliability at scale. The companies that sustain advantage are those that embed into core business processes, particularly in areas such as AI SaaS, AI/MLOps, Software development lifecycle, infrastructure, and enterprise workflows.
In contrast, emerging fields like quantum computing, quantum communication, and quantum sensing are still at the capability-building stage, where the primary question is not value capture, but achieving stable, scalable breakthroughs and integrations in quantum systems.
.svg)
.svg)
.svg)
.png)
.png)
.svg)
