Summary of Minimum viable (data) product

00:00:00 - 00:30:00

The video discusses the concept of the minimum viable product, or "data product," and how it can be helpful in developing software products. The data team at Slido emphasizes the importance of learning from each other and collaborating with other departments within the company. One of the challenges faced by the data team is the fragmentation of the development workflow and the all-up cube problem. The video discusses the process of creating a minimal viable data product, focusing on the challenges of dealing with data that is fragmented and constantly evolving.

• 00:00:00 This session is discussing how to bring a product mindset to analytics engineering, with particular emphasis on the use of Slido's semantic layer. Slido provides an approach to analytics engineering that incorporates product development best practices, and they will be presenting their workflow to illustrate this.
• 00:05:00 The video discusses the concept of the minimum viable product, or "data product," and how it can be helpful in developing software products. The data team at Slido emphasizes the importance of learning from each other and collaborating with other departments within the company. One of the challenges faced by the data team is the fragmentation of the development workflow and the all-up cube problem.
• 00:10:00 The video discusses the process of creating a minimal viable data product, focusing on the challenges of dealing with data that is fragmented and constantly evolving. The presenter explains that the fragmented development workflow is a problem because it makes it difficult to track and manage data. To overcome this, they create an all-up cube that contains 200 metrics. This allows the business user to see the state of the product at a glance.
• 00:15:00 The presenter discusses the benefits of the DBT semantic layer, which solves issues with the fragmented workflow and lack of version control. They discuss how Deep Note is a collaborative notebook tool that allows data scientists to work with their team in real time.
• 00:20:00 The video demonstrates how a team used deep note, a deep semantic layer for data product development, to codify and process key metrics from an op cube. The team also used the DBT Cloud engine to test and run macros, and integrated the data product with other downstream tools. Finally, they built a notebook to serve as a documentation and exploration tool for the data product.
• 00:25:00 The video discusses how the brand new integration with the DBT semantic layer has made it possible for data engineers to prototype against a DBT server, write code, and debug all in the same interface. The impact of this is that speedometric development goes down, and analysts are now focused on what they're great at - allowing teams to self-serve and be more self-reliant. This fosters data literacy in the broader team and helps analysts and business users focus on what they do best - knowing their tables and applying their domain knowledge.
• 00:30:00 The presenter discusses Minimum Viable Products, which are a way to test various aspects of a product before launching it to the general public. He also talks about how transformation can help with data management, and how Deep Note can help with data engineering.