This week on The Data Stack Show, Eric and Kostas chat with Rishabh Bhargava, Co-Founder and CEO of refuel. During the episode, the group discusses the evolution of AI, machine learning, and large language models (LLMs). Rish shares his background and the inception of refuel, which focuses on making clean and reliable data accessible for businesses through data cleaning, labeling, and enrichment using LLMs. The conversation explores the impact of LLMs on data quality, the challenges of implementing LLM technology, and the user experience of working with LLMs. They also touch upon the importance of confidence scores in machine learning and the iterative process of model training, a practical use case involving refuel and RudderStack, and more.

This week on The Data Stack Show, Eric and Kostas chat with Viren Baraiya, the Co-Founder and CTO of orkes.io. During the episode, Viren discusses the evolution of orchestration in the context of AI and large-scale systems. The group discusses the transition from Viren’s work at Netflix to founding orkes, the challenges of integrating AI into applications, and the importance of orchestration to manage these complexities. He also highlights the non-deterministic nature of AI, the need for guardrails, and the potential for AI to change technology interaction. The episode also covers the recent move of Netflix’s Conductor project to a community foundation, the future of AI in business and its impact on job creation, and more.

This week on The Data Stack Show, Eric and Kostas chat with a panel of experts as Wes McKinnyey (Cofounder, Voltron), Ryan Blue (Co-Founder and CEO, Tabular), Chris Riccomini (Seed Investor, Various Startups), Pedro Pedreira (Software Engineer, Meta), all share their thoughts around the topic of composable data stacks. During the conversation, the group chats about the importance of open standards and APIs for efficient interoperability in data management systems, the evolution of data workloads, the need for specialization, and the challenges in building composable components. The conversation also covered the significance of an intermediate representation (IR) for decoupling various layers of data systems, the complexities of data types, and the desire for more secure data sharing methods. The panelists explored the evolution of open standards and the trade-offs between composable and monolithic systems, expressing excitement about new data infrastructure projects and technologies, modular execution engines, new query interfaces, standardizing policy decisions across different data management platforms, and more.