Gaining a 360-degree View of the Customer
3.12.2020Industry: Insurance
Modra Zavarovalnica d.d. is a Slovenian supplementary insurance company selling additional pension insurance. Their selection of products provides increased social security to individuals and their families during all stages of life as well as helping them maintain their lifestyle after retirement. Migrating to Snowflake cloud data platform allowed them to gain a unified view of the customer with all of their activities present, while also significantly decreasing the total costs of ownership.
The Challenge
The company recognized the need to renovate their existing data warehouse because of its limited amount of data space and slow processing speed resulting in resource contention. Existing on-premises data warehouse built on MS SQL was running out of resources (storage and compute power) and therefore the load process, even though heavily optimized, took more than 8 hours on a daily basis. The second challenge was the integration of web sources like Sales force and Mailchimp into the existing environment.
They wanted to organize their content by areas of assets, liabilities and costs because they needed an in-depth view of their customers. Modra wanted to establish a comprehensive data warehouse for all their data sources (back-end systems, cloud services, CRM, Mailchimp and other structured and unstructured data content). They saw their future central data warehouse as a source for development new products, new applications and refine services based on advanced analytics in terms of mathematical algorithms for data mining or data science (probabilities of departure, advanced segmentations, probability of deposit, next best offer algorithms, improvement of user experience via digital channels, etc.).
Proposed Architecture
Implementing a modern data warehouse running on Snowflake cloud data platform brings better insights for each and every customer and provides an exceptional service due to no limitations on compute power and storage, per-second billing and near-zero maintenance. By modernizing the existing data warehouse and moving from an on-premise Microsoft SQL to the Cloud data platform, Snowflake enabled the client to start building a unified data warehouse with additional aforementioned web sources, web browsing history and mobile interactions.
During the first step, In516ht moved all the data from the web sources such as Salesforce, and Mailchimp to Snowflake data lake which took two days. In the second step In516ht moved on premise transactional database running on Oracle to the Snowflake. All the workloads were moved to the Snowflake using the antFarm data movement tool and daily uploads of newly generated transactional were set up.
With DataMerlin, a meta-data based data warehouse automation tool, In516ht automated every step of the data warehouse integration process. They switched from MS SSIS traditional ETL based approach to modern ELT using DataMerlin to take advantage of cloud compute resources. Two main advantages were realized in the process, client set the ground rules for modern ELT development and the development time was decreased by 60%.
Snowflake and Data Merlin, among other internal processes that were introduced by Modra, enabled the company to comply with GDPR requirements. Data Merlin provides full data lineage, so analysts can now track data from Snowflake back to its source.
Analysts can now address all new use cases and the simultaneous operation of various activities without affecting one process to another (e.g. analysis marketing data does not affect the rapid production of predictive models). Since no administration is required by the back-office IT (Snowflake operates as a cloud service) it is also very affordable.
Outcome
The aim of the project was to create a central data lake repository on a modern platform that will allow the processing of various amounts of data as well as provide minimal administration for data warehouse maintenance (data warehouse as a service). With the new platform, business users can create tailor-made campaigns for different client segments and analyse all the data in one place. Additionally, finance analysts can analyse the investment portfolio for a specific financial instrument and have a daily snapshot of data and different performance analysis. ETL running time is reduced from 8 hours to 2 hours.
Client could finally move away from traditional Excel-based reporting, where they were struggling with visualizations due to Excel limitations and ensured all departments are looking at the same picture.
Tableau was chosen for BI analytics because it enables the customer fast adoption and amazing visualizations. Advanced visualizations help customers focus on visual analytics rather than the old-fashioned grid reporting in order to quickly realize deviations of their business on an operational and strategic level. Furthermore, they utilized Tableau server infrastructure where they publish and store operational and management dashboards and use Tableau subscriptions to automate Dashboard delivery to key stakeholders.
Customer Statement
Choosing the In516ht service provider to deliver an end 2 end data analytics solution was a good decision. Their knowledge and understanding of the financial industry were crucial for successful project delivery because we were able to focus on use cases and not on explaining our business process and terminology during the data preparation phase. With the help of In516ht, we analysed several different technology approaches but in the end decided to go forward with the future proved cloud data platform Snowflake. Simply said, it’s an amazing technology. We were able to significantly cut down operational costs, because Snowflake charges only for what you use, there is no need for maintenance and most importantly, we are now able to address any use case because we have virtually unlimited resources. Snowflake works marvellously with Tableau BI tool, which we have chosen to bring visualizations to the next level. It enables us to prepare sophisticated dashboard visualizations and it quickly points out deviations in our daily business.
Ernest Dulič, IT Deputy Executive Director
