Data-powered startup ecosystem: Building a common data-platform for The Mobility Ecosystem of Baloise

The Mobility Ecosystem of Baloise wants to accelerate the growth of the startups in its ecosystem and help them to become data masters which are able to utilize data and AI in their business. As a first step in this journey, The Mobility Ecosystem of Baloise worked with Capgemini to build a common cloud-based data platform for the ecosystem that enables the startups to gain insights for the ecosystem that enables the startups to gain insights, create new data-powered services and optimize business models.

Benefits for the customer in a nutshell

• Modern data platform and architecture available for all current and future startups in the ecosystem
• Enabled data-powered business optimization and new service creation
• Easy onboarding and access to the data platform and a data ecosystem for new companies

Starting point

The Mobility Ecosystem of Baloise for startups operating in the general field of mobility: How do we move from A to B in a smart, safe and sustainable way?  Their business includes shared vehicle operation, services to fleets, alternative ownership models, and other consumer-friendly services.

Investing in modern technology and data-powered solutions is often difficult for companies in their early stages. It requires understanding of opportunities and resources – both financial and human. Technologies might be too complex for small-scale companies and recruiting a team of data scientists and architects to build a solution is both demanding and expensive.

At the same time, the possibility to use data can make processes more efficient, bring clarity to business and accelerate growth with data-powered products, and services that create a lasting competitive edge.

The Mobility Ecosystem of Baloise wanted to help their ecosystem of startups in becoming more data-driven and decided to offer a common data-platform that would allow the companies to gain all these benefits from the very beginning of their journey.

The Mobility Ecosystem of Baloise chose to collaborate with Capgemini in building the data platform and showcasing the possibilities via selected use cases as they valued Capgemini’s capabilities in modern technologies and experience in the mobility industry. They also felt that the team understood the objectives and agile approach very well.

We considered Capgemini as an interesting partner to collaborate with because of their previous experience in the mobility industry.

Sami Pippuri, CTO in Residence, The Mobility Ecosystem of Baloise

Requirements for a shared data-platform

While building a data platform for an ecosystem of startups, there were several requirements that needed to be considered.

First, there were a set of expectations for the architecture from the technology point of view. It was crucial for The Mobility Ecosystem of Baloise that the architecture would be a modern, scalable cloud solution providing a solid ground for experimenting and implementing new applications & services. Additionally, the platform should be able to ingest and store data from various sources, both internal and external. Moreover, as there were several players in the ecosystem with their own technical set-up, it had to be flexible for multi-cloud environments. Not to forget the analytics and AI capabilities of the platform.

Secondly, data privacy and security were crucial. The data should be securely stored with only the correct parties having access to the data. In practise, this required isolated instances/accounts as several startups would be using the shared data platform, while maintaining full isolation of the data stored within.

Thirdly, maintenance and administration should be minimal and the learning curve as flat as possible. While The Mobility Ecosystem of Baloise is offering the data platform to start with, ideally over time, the portfolio companies would take more responsibility over the assets. A desired future would be to have an established “data community” where portfolio teams can help each other, sharing resources where possible.

Iterative process in collaboration with chosen startups

Capgemini and The Mobility Ecosystem of Baloise started building the platform in an agile and iterative way, with a hypothesis to prove. The project consisted of three phases:

1. Design & specifications: In the first phase, the basic requirements for the data platform were discussed and the best-fitting technologies catering to needs were selected. Part of this process was also to understand the actual use cases that the startups might have. These insights were gathered in workshops and numerous discussions between the project team and key stakeholders from The Mobility Ecosystem of Baloise and startups.

The chosen core technology was Snowflake which met all the requirements listed in the previous section, from multi-cloud support to ease of maintenance.

The use cases for further development were related to car sharing business both from renting and fleet serving perspectives.

2. Implementation of the data platforms and pipelines: A Snowflake data platform was set up in such a way that the master account was located on AWS and the separate associate accounts for the startups could be hosted on any public cloud – the The Mobility Ecosystem of Baloise portfolio companies are typically on AWS or Azure. The target portfolio companies were on AWS. Data pipelines first transferred the raw data to data lake (AWS S3) from where it was then automatically moved to Snowflake via Snowpipe.

3. Building of analytics & machine learning (ML) solutions: The studied use cases tackled different aspects of car sharing business from demand forecasts to optimizing recharging and relocations services. The solutions were built utilizing Snowpark framework and results visualized both in Jupyter notebooks and selected BI tools including PowerBI and Looker.

Achieved outcomes

The data platform is now up and running. The start-ups which were involved in the project can continue creating new insights and optimize their business based on the existing data sources or creating additional data integrations themselves. New companies can be easily onboarded, and they can start building basis for their own data mastery. And all this is done according to the highest principles of data privacy and security.

The portfolio teams were impressed by the capabilities that the data platform could present in only a short period of time, thanks to using the modern cloud based tooling, reuse of the platform from project to another and the high level of expertise in the project.

The machine learning (ML) models created for forecasting business outcomes and optimization of EV charging for example, were beyond the initial guesstimates where we might get to in the use cases, within the boundaries of the project budget. Yes, the project concluded within budget.

Experimenting with various use cases proved the value of having a common scalable data platform and showcased possible ways of building advanced analytics on top the platform.

For example, if you are running a car sharing business, you need to understand which car locations make the best sales figures efficiently. The external data you want to use can be for example weather forecasts and if there are any special holidays or events in the calendar that would typically increase the need for rented cars.

The next step is to combine with internal data to understand which zip codes are best sellers at the considered time. Based on this information, you can move the cars to the right locations and make also sure that they are sufficiently charged for the potential clients.

Petteri Mannersalo, Principal AI Architect, Capgemini Finland

Experiences from working with Capgemini

The Mobility Ecosystem of Baloise chose Capgemini as a partner because of the global capabilities for different data platform solutions. As a technology-agnostic company it is capable of choosing the best possible solution for each need.

The Mobility Ecosystem of Baloise enjoyed working with a reliable partner like Capgemini and is happy with the outcomes of the project. The combination of well- structured project management, with clear milestones, and professional team members were essential in successfully executing the project.