The CEO Behind the Fastest-Growing AI Inference Company | Tuhin Srivastava - Gradient Dissent: Conversations on AI Recap

Podcast: Gradient Dissent: Conversations on AI

Published: 2025-11-18

Duration: 59 min

Summary

Tuhin Srivastava shares his journey with Base10, a rapidly growing AI inference company, emphasizing the importance of perseverance and adaptability in the ever-evolving machine learning landscape. He discusses the transition from model serving to inference and the role of infrastructure in scaling AI applications.

What Happened

In this episode, host Lucas B. Wald engages with Tuhin Srivastava, the CEO and founder of Base10, a company that has emerged as a leader in the AI inference market. Tuhin reflects on his extensive experience in early-stage companies since 2011, detailing the struggles and learning curves he faced before Base10's recent success. He highlights that despite the perception of rapid growth, there were years of uncertainty and hard work that preceded their current momentum, which finally surged due to market evolution and the arrival of significant opportunities in the AI field.

Tuhin elaborates on the transformation of the AI landscape over the last few years, particularly the shift from small models for internal use to large models requiring robust infrastructure for production. He points out that while model serving was once considered a commodity, the demands of scaling and reliability have increased significantly. This shift prompted Base10 to pivot towards a more dynamic approach, focusing on the infrastructure needed to support these larger models effectively. Tuhin emphasizes the importance of not giving up and maintaining a cohesive team during challenging times, which ultimately led to their breakthrough.

Key Insights

Key Questions Answered

What led to Base10's growth in the AI inference market?

Tuhin Srivastava explains that Base10 experienced a significant shift in its trajectory due to changes in the AI market, particularly the evolution from small models for internal use to larger models requiring robust infrastructure. The recognition that model serving was not merely a commodity but a critical component of scalable AI applications helped position Base10 as a leader in the inference space.

How did Tuhin Srivastava's past experiences shape Base10?

Tuhin shares that he has been involved with early-stage companies since 2011, experiencing numerous challenges and failures before founding Base10 in 2019. These experiences taught him the value of persistence and adaptability, which were critical in navigating Base10 through its initial struggles towards becoming a successful enterprise.

What factors contributed to the pivot in Base10's strategy?

The pivot in Base10's strategy was influenced by the changing landscape of AI, where the company transitioned from targeting data scientists with small models to addressing the needs of larger, production-scale models. This shift underscored the importance of infrastructure and reliability, which Base10 had already been developing, allowing them to adapt without a complete overhaul.

What advice does Tuhin offer to entrepreneurs facing challenges?

Tuhin emphasizes the importance of not giving up, working with a supportive team, and being responsive to market changes. He suggests that finding a balance between perseverance and adaptability is key to overcoming obstacles, as it was for Base10 during its early years when growth was not yet evident.

How does Base10 differentiate itself in the AI inference space?

Base10 differentiates itself by focusing on infrastructure and scalability, which are essential for deploying large AI models in production environments. Tuhin notes that while many companies might view model serving as a commodity, Base10 understands the complexities involved and addresses them directly, positioning themselves as a reliable partner in the fast-evolving AI landscape.