The Analytics Engineering Podcast

Journalist/data practitioner Matthew Lynley discusses the rise of Gen AI in the media, his journey from mathematics to journalism, encountering Chat GPT, challenges of monetizing an AI newsletter, AI content creation, and the convergence of fields in AI development.

Topics discussed include semantics, text-to-SQL performance, AI's impact on structured data analytics, evolution of semantic web, knowledge graphs, observability in business analysis, GPT-4 for text to SQL, design patterns in LLM interaction, and knowledge management in the data industry.

Exploring the evolution of the modern data stack, AI integration in analytics, Mode's transformation to a cloud-first company, AI-powered formula builder for SQL queries, redefining the modern data stack, empowering business through IT and data teams, and envisioning the future of the data industry.

Moritz Heimpel from Siemens and Ben Flusberg from Cox Automotive discuss their experiences adopting a data mesh architecture and collaborating with data at scale in large organizations. They explore topics such as setting up infrastructure to scale, domain specific ownership of data, data security, and the future of the data industry.

Brandon Taylor and Michael Olschimke discuss the Data Vault approach, its adoption in Europe, alignment with data mesh architecture, and the ongoing debate over Data Vault vs. Kimball methods. They also explore the impact of DataMesh on data warehouse design, recommended data modeling approaches, and hopes for the data industry.

Jonathan Frankle, Chief Scientist at MosaicML, discusses the future of training specialized models, MosaicML inside Databricks, and responsible AI practices. They explore LLM-based systems, one-way hash functions, and the integration of Databricks platform. The trade-off between model size, cost, and complexity in AI training is also touched upon.

Kasey Mazza, analytics engineering manager at HubSpot, discusses the roles of data analysts and analytics engineers, building internal data communities, and the evolving landscape of data teams. They also explore career growth and satisfaction, the interplay between organizational structures and tooling, challenges in decision-making and central governance, using DBT across the company, and the future of the data industry.

Arjun Narayan, CEO of Materialize, and Nathan Bean, a data leader at General Mills, discuss operationalizing warehouses, streaming analytics, and the challenges of manufacturing cereal. They cover the maturation of streaming technology, data management challenges, real-time operational decision-making, managing variation in manufacturing, digital twins for manufacturing line automation, operationalizing warehouses, trade-offs between batch and real-time analytics, the evolution of streaming analytics, and query languages in data analytics.

Yannick Misteli is the head of engineering for the go-to-market domain at Roche, a $250 billion multinational pharmaceutical and diagnostics company.  Roche was an early supporter of dbt Cloud, and Yannick helped move his team of 120+ engineers to a modern data stack. He always finds a way to push the boundaries to make a large company founded in 1896 incredibly modern and innovative. We wanted to know more about the "how" of the work—the people, process, and technology.  Read more about Roche's data journey here: https://docs.getdbt.com/blog/dbt-squared

Andy Pavlo, a professor of databaseology at Carnegie Mellon and founder of OtterTune, talks about the complexity and specialization of database systems, the trend of separating storage and compute, scaling challenges, and the future of the data industry.