Shiva’s $10M Fund: A New 'Tiny Team' Blueprint for Brazilian AI

The launch of Shiva, a $10 million Brazilian investment fund, marks a significant shift in the venture capital landscape, specifically tailored for the era of generative AI. By focusing on "tiny teams" of one to three individuals—whom the fund calls "Stars"—Shiva is betting that the increased productivity afforded by AI tools makes massive headcounts and multi-million dollar seed rounds unnecessary for early-stage software companies. This model, led by Lucas Marques, a former partner at the publicly traded fintech Méliuz, suggests a move away from the traditional Silicon Valley "blitzscaling" approach toward a more capital-efficient, sustainable growth trajectory.

Central to Shiva’s strategy is a unique funding structure. Instead of the typical large, upfront equity checks that often lead to premature scaling and high burn rates, Shiva provides monthly stipends to founders for up to a year. Total funding per company is capped at $300,000, and the fund’s equity stake is capped at 15%. This "lean" methodology is designed to support niche software companies that can reach profitability quickly. The "Stars" designation refers to a specific type of founder: highly technical, resourceful, and capable of leveraging Large Language Models (LLMs) to perform the work of an entire department. In Shiva's view, the modern AI stack effectively replaces the need for a traditional C-suite in the earliest stages, allowing the fund to focus its resources on individual vision and execution speed.

“

Total funding per company is capped at $300,000, and the fund’s equity stake is capped at 15%.

The fund’s mission is deeply rooted in social impact and the personal history of its founder. Marques, who grew up in rural Brazil, previously founded the NGO Programadores do Amanhã (Tomorrow’s Programmers) to teach coding to low-income students. Shiva is an extension of this work, aiming to democratize entrepreneurship by lowering the financial barriers to entry. In an environment where AI can automate significant portions of development, marketing, and administration, Marques believes that founders from disadvantaged backgrounds can now compete globally without needing the massive capital reserves typically required to hire large engineering teams. By keeping overhead low, these startups can achieve life-changing outcomes for their founders through mid-sized exits in the $20 million to $50 million range—a segment of the market often ignored by larger VC firms chasing billion-dollar valuations.