Data Breakfast at T-Bank in Early January 2025 year (Data column)

I was finally able to see the reports from breakfastThat was almost a month ago. It was performed by Dima Anoshin, author of the channel "Data Engineering" (@rockyourdata)As well as Valera Polyakov, our data director. I was a bit involved in organizing the event, but I couldn't attend because I was flying home from Sri Lanka. And the topics were interesting: Dima talked about how data projects look in Western companies and what technologies they are built on, because Dima has experience in Amazon, Microsoft and a number of smaller companies. Valera talked about the evolution of T-Bank’s approaches to working with data from the beginning of time to the present moment.

Let me tell you briefly about each of the reports.

1) Modern cloud solutions for analytics and their importance for business from Dima Anoshin Dima began with a story about the importance of analytics to increase profits and reduce costs. Next, he talked about conceptual analytical solutions, which basically consist of sources, storage and processing systems.

• Dima shared. 11 Real projects he has been involved in over the past ten years. For these projects, trends were clearly visible - the transition from on-prem to cloud, the transition from all-in-one to separate storage and compute, the rise of cloud analytical solutions snowflake and databricks, the design of the role of data engineer as people who do infrastruct under dataops.

2) The history of the data platform in T: from SAS to PaaS from Valera Polyakov History of data infrastructure development with 2007 years. **- The SAS era. (2007 - 2011)**Work with proprietary SAS tools before switching to Greenplum. **- The Greenplum era (2012 - 2016)**Moving to MPP databases, selecting Greenplum, scaling infrastructure. There's still a wrong turn with SAP BO in 2014 year:) - An era of increasing complexity (2017 - 2021): Here, the company was actively growing and the scaling limit for a single Greenplum cluster was reached. Further growth was through the introduction of multicluster and proprietary solutions for data replication. This has led to a rapid increase in load and complexity with data management. **- Age of change. (2022 - 2026)**The era of change that brought modern approaches Democratization of data engineering and platformization. Transition to cloud nativeity of the data platform using a multitenant architecture. Implementation of the concept of “data as a product” with an emphasis on quality metrics. Change brings new challenges that we are working on right now. Standardization of data between domains. The importance of centralized practices for data methodology. Balance between tactical objectives and strategic development.

In general, the data breakfast turned out to be excellent, the reports were dense – I hope that this will become a tradition and my colleagues from the data platform will delight us with interesting and useful events more than once.

P.S. Late last year, I told A report on the evolution of architecture in T, which is very similar in structure and logic of the narrative to the Valerin report. I ended up catching déjà vu moments as I listened to this talk about the evolution of the data platform:)

#Engineering #Data #Architecture #Storytelling