Indian artificial intelligence startup founders are significantly more capital-efficient than their US counterparts, according to Nitin Kaimal, partner and chief operating officer at Bessemer Venture Partners, as the firm sharpens its focus on early-stage AI and SaaS bets.
The Silicon Valley-headquartered investor, which manages around $20 billion globally, typically invests at the seed to Series B stages and backs one or two companies in a category rather than spreading capital across multiple small bets.
“Seed rounds and Series A or Series B is a sweet spot for us. We don’t like to do a lot of small deals,” Kaimal told FE in an interaction, adding that building category-defining companies requires time and concentrated support.
In India, Bessemer has backed companies such as Swiggy, BigBasket and Urban Company. It is now evaluating AI-led SaaS startups that it believes can challenge incumbent IT services firms.
Agility Edge
Kaimal said that traditional IT majors such as Infosys, Wipro and Tata Consultancy Services are constrained by labour-arbitrage models in an AI-led environment.
“Why would a client want to sign a $2 million contract with 400 people when SaaS startups can deliver the same solution with 100 people and AI software tools?,” he said.
According to him, projects that may take eight to twelve months at a traditional IT firm can be executed in six to eight weeks by AI-native startups.
Bessemer recently participated, alongside Sorin Investments, in a $4.6 million seed round in Aivar AI, founded by former AWS engineers.
The company has about 80 enterprise customers, Kaimal said.
He added that the firm evaluates startups primarily on measurable outcomes. Companies must demonstrate a step-change in a cost, inventory or customer conversion metric rather than incremental gains.
“Most of these startups are generating a sizeable annual recurring revenue already. So, we speak with five of their customers to gauge their reaction,” he said, adding that strong customer validation is essential.
Beyond the ‘Wrapper’
Kaimal also said the phase of simple GPT-wrapper startups that emerged after the launch of OpenAI’s ChatGPT in 2023 is largely over.
Sustainable differentiation, he said, requires combining large language models with domain context and advanced prompt engineering to create distinct intellectual property.
Comparing ecosystems, he said the US has more research-led AI firms, while India is services-oriented. However, he described Indian vertical AI startups as more frugal, with a sharper focus on cash flow and cost discipline rather than aggressive marketing spends.