More with less: How AI startups are scaling rapidly with lean teams, high revenues

Representative image

At a time when enterprises are trying to cut costs and trim headcount, a new generation of AI (Artificial Intelligence)-native startups are scaling rapidly with leaner teams.

For example, San Francisco-based AI code editor Cursor built by a team of just 20, has reportedly clocked $300 million in annual recurring revenue (ARR) in less than a year. AI-powered app builder Lovable, headquartered in Sweden, reached $40 million in ARR within five months, with a team of only 15.

Indian companies too are not behind with the likes of Sarvam AI, Karya AI, Avaamo AI, and CoRover.

“Gen AI startups are at a vantage position to build, scale, and augment AI solutions for the enterprises on the go and with time truncation unlike the AI firms of the past, these startups have created robust low code-no code Gen AI platform,” Sameer Dhanrajani, CEO of 3AI, told Moneycontrol.

Earlier this month, responding to a question on what the future of software engineering jobs look like in the AI-era, Perplexity founder Aravind Srinivas told Moneycontrol that he sees endless possibilities with AI but not for those who don't adapt.

The AI-powered answer engine founder said AI tools now help coders to launch and grow products independently, and even build million-dollar businesses with just one or two people.

Going back to the boom in digital demand post-pandemic, the trend back then showed that cross-functional and nimble teams can be just as effective. Moreover, with companies doubling down on talent-based hiring and reducing operational costs, hyper productivity per employee is a key metric now.

Effect on jobs

In an AI-driven world where machines can code, automate, and optimise with speed, the future of the human workforce will depend on on strategic reskilling, driven not by corporate mandates but by personal ownership.

"The professionals who will thrive in the AI age will be those who take control of their learning beyond corporate-sponsored courses. They will need to ask themselves: Do I want to be an AI creator or an AI operator?" says Rituparna Chakraborty, cofounder, TeamLease Services.

As AI automates routine roles, mid-level professionals managing programming and infrastructure face the greatest disruption.

To stay relevant, they must go beyond company upskilling and choose whether to become AI creators, those who design and build intelligent systems, or remain operators.

Modus operandi

“The maturity and accessibility of AI tools have made this transition feasible. What was once the domain of tech giants is now open to startups and enterprises alike,” Soumendra Mohanty, Chief Strategy Officer at AI solutions firm Tredence.

New-age Gen AI startups are able to build sophisticated algorithms and models using retrieval-augmented generation (RAG) techniques without needing massive AI teams, Dhanrajani adds.

RAG is a technique where AI pulls in external information on-the-go, to give smarter and more accurate answers. This eliminates the need for large in-house teams to build and maintain extensive proprietary datasets.

Dima Shvets, a scout with Silicon Valley venture capital firm Andreessen Horowitz, believes that with next-gen AI tools, a few exceptional people can build software-centric businesses that scale to over $100 million in revenue, on the back of automation and not headcount.

The Dream Team

A high-performing AI-native startup team would typically include three core roles, according to Shvets.

Firstly, a business visionary who masters go-to-market strategy and quickly validates ideas with users and partners.

Second, a tech or product visionary who anticipates major technology shifts and thinks at the infrastructure level.

And lastly, a systems builder, whether a CTO, marketer, or operations expert, who can bring structure and scale to fast-moving chaos.

In fact, Tredence has already set up a studio, where small pods use extreme programming and ultra-agile methodologies to develop go-to-market solutions in just a few weeks.

“Lower burn, faster iteration, less dilution, more fun,” tech investor Shvets wrote in a professional networking site.