Toyota Motors Europe is bringing all its enterprise information together on a Snowflake platform that will provide a foundation for well-governed data access.
The project is all about providing better access to information. The aim is to move away from a siloed approach to data, where each business unit holds its own information, to a cross-organization approach, which supports more effective analytics processes. Thierry Martin, Senior Manager for Data and Analytics strategy at Toyota Motors Europe, explains:
It's bringing a solution to data, which is scattered everywhere, and it’s about quick access to data. We recognised we needed to allow people to access data much more easily and that's why I was looking for a cloud-based data warehouse solution. There were different solutions, of course. But very quickly, we narrowed on Snowflake.
Toyota Europe started using Snowflake in May this year. After looking at the market, the car giant created a first use case to trial the technology, which involved setting up a development environment. IT professionals used the trial to test the platform and Martin’s team started to define the specifications for an organization-wide data mesh.
The first project was a success and his team produced a 50-page handbook that outlines ways of working:
To write that document, we needed some time to understand what we are working with. We wanted to have the perfect integration that allows us to bring in data from other technologies. Snowflake is one of the few platforms that allows you to do that.
Right now, the enterprise-wide solution is not in production. The key aim is to learn more about how Snowflake works, but there are big aims for the project through 2024, says Martin:
We want to use Snowflake as the central piece of the puzzle where everything is connected. And with Snowflake Professional Services, we have refined the blueprint for our data mesh. The first big use case for that will be around delivering business intelligence to the business.
Moving into gear
The Snowflake platform will enter production towards the end of this year, using external integrators who will help his team build the first level of enterprise-wide data integration. As part of that process, data will be moved onto the Snowflake platform. Once the first stage is complete, Martin is looking to expand the effort quickly and build the data mesh:
The idea is to create one first domain, where we are developing products and everything is under control. But the plan is to distribute planned domains very quickly that we provide to logistics, quality or R&D. This is where we have the data mesh and we’re doing that through a governance model. This governance will help the business to upload their data products into the data marketplace.
Martin defines Toyota Europe’s data mesh as an approach to governance that ensures responsibility for data products stays with the business owners. It’s these people, he notes, who have in-depth knowledge of enterprise data:
It's about ensuring that we have the data owners, data stewards and all these roles, which are super-important. When new data products are proposed, the first question to the business will be, ‘Who is the data owner?’ And then we will work with the data-engineering team to move the data into Snowflake.
Toyota Europe’s data mesh will also draw on a range of other technologies. The company will use Dataiku for collaboration, Collibra for governance, and Denodo to connect data meshes across different parts of the organisation, such as Toyota Europe and Japan. Data owners won’t get to see Snowflake in action. Instead the data platform will work as a back-end platform behind a data marketplace that allows people across the business to use the right kind of data for their projects. Martin explains:
If our internal customers are interested, they can browse the data marketplace and then request access to data products. By doing that, the workflow is automatically started, so the owner will receive the request to provide access to the data.
The long-term aim is to ensure the production version of the Snowflake implementation gives employees quick access to consolidated data sources that might once have been scattered around the organisation in siloes. He says senior executives within Toyota Europe have already seen the potential long-term benefits of the approach:
When I show the business what Snowflake can do – such as the Marketplace in Snowflake and how it's so easy to get access to data from outside – they’re impressed. And then I show them how we will have the same process for internal data; that they will just get access to data and it will be available immediately. Then, I can see their eyes opening.
As someone who’s currently in the middle of a Snowflake implementation, Martin advises other business and digital leaders who are thinking about going down the Snowflake route to learn about how the technology works. He’s writing Python code, he’s adding to his SQL knowledge, and he can operate Snowflake. He says other IT leaders should do the same:
I think it's very important for the leader to be able to not just understand at a high level what the platform is, and then provide a budget for the activity, but to also understand the technology in more detail and its limitations, especially given what's coming next with generative AI, chatbots and all these other emerging technologies.