How Floryn makes faster decisions based on data with Deepnote

One of the main advantages of using Floryn as a lender over traditional banking is the speed at which a loan can be arranged. To be able to arrange a loan within 24 hours for its customers, Floryn relies on machine learning models that go through large volumes of bank transactions every day to assess an applicant's eligibility. The volume and implications of potential loan approval both place an immense importance on the precision of such models.

The data science team works in short cycles of a few weeks to ship improvements to its models or build new ones. They needed a platform where they could work efficiently through the whole lifecycle of a model, from exploratory analysis, experimenting and building the model to finally  shipping it to production.

Working with data often comes with additional security needs, and working in the financial industry comes with very strict compliance requirements. Floryn is no exception to this, since they are regulated by entities like the Dutch Central Bank and work with financial transactions data. Given their need to adhere to the highest security standards, they needed a trusted partner to provide a solution that handles their data with certified security, privacy and compliance practices.

On the data analytics side, analysts were used to working with BI tools like Looker and working directly with stakeholders across different teams. Looker is a great tool for viewing finalized dashboards, but there was no easy way to collaborate efficiently with their many stakeholders., with each exploratory question turning into a dashboard that analysts would have to spend hours rebuilding with each subsequent follow-on question.

With these challenges in mind, Floryn chose Deepnote as their trusted partner.

Deepnote's collaborative and interactive platform enhanced Floryn’s machine learning workflows through real-time collaboration, interactive coding, data visualization, integrations, and scalable compute resources. Its user-friendly interface and Python library support streamlined the process of modeling data, making it an ideal choice for efficient and collaborative machine learning projects. Making insights accessible to stakeholders has become more straightforward than ever before with Deepnote.

Security was a key requirement for Floryn’s data team. Deepnote prioritizes security by implementing robust data encryption, access controls, and authentication mechanisms to safeguard user data. Its secure collaboration features allowed Thom and his team to work together confidently while maintaining data integrity.

Floryn’s data analytics workflows especially benefited from Deepnote’s collaborative features. Analysts saw an opportunity to shorten the feedback loops when working with stakeholders by involving them early on in the data exploration stage before going to Looker, saving many hours of back and forth communication.

With Deepnote, Floryn’s data science and analytics teams were able to improve efficiency and meet their demanding timelines, saving time by either squeezing in more value in each cycle, or by shortening the feedback loops with stakeholders. Beyond the data team, Deepnote enabled Floryn to put data at the center of the organization, reaching every team and empowering them to make faster data-driven decisions.