Skip links
Streamlining data extraction and analytics to reduce operating costs and raise

Streamlining data extraction and analytics to reduce operating costs and raise analytics accuracy

Streamlining data extraction and analytics to reduce operating costs and raise

Streamlining data extraction and analytics to reduce operating costs and raise analytics accuracy

The Client:

A national-level financial services player, part of a legacy conglomerate active in several business areas, including consumer goods, appliances, agri-based businesses and real estate. The client’s business strategy is aligned with the nation’s aspirations, in that it is committed to providing home loans across national demographics, enabling their customers to realize their home ownership dreams and the nation’s mission to raise the standard of living.

The Need

The company urgently needed actionable, accurate date to make business decisions. In this they were hampered by their legacy system that used an Online Transaction Processing (OLTP) platform to capture, store, and process data from transactions in real time. The company was using PostgreSQL as the database for OLAP workloads. However, as data was being stored in different tables, the customer was forced to expend considerable effort to extract the reports needed to generate data required for various decisions. The lack of a data warehouse in the legacy system further complicated analytics. The client needed a comprehensive data analytics platform that would enable higher degree of data engineering, streamline data processing, automate reporting and enable extraction of usable business insights.

The Solution

Teleglobal drew up a solution that combines AWS services and Snowflake to extract data from the client’s existing platform and pipe it into a OLAP platform so as to enable immediate data modelling. This would meet a number of customer objectives.

Business data was gathered in real time and stored on the OLTP platform before being extracted into AWS S3, which would provide both the raw layer and cleansed data. Any future ML requirements would be taken care of using AWS services

The cleansed data was then pulled into Snowflake’s published layer. Data modelling would finally be performed in the published layer (in Snowflake)

Benefits

⦿ Teleglobal’s solution provided the client with a single source of truth for their data—unstructured, semi-structured, and fully structured

⦿ Enabled creation of tables and drawing the data into a single platform

⦿ Highly secure platform, requiring stringent authentication to access

⦿ Allowed the client to integrate reporting tools as per their needs, so as to :

– Perform gap analysis

– Build fully configurable ETL pipelines (AWS to Snowflake); addition of new

– Automate population of certain fields in business operation reports or related key performance indicators, viz:

–  Financial Reporting: delinquency rate, interest spread Profit before tax, value     insurance, premium, penetration

–  Loans and Collections: total number of loans, amounts, ROI, EMI, outstanding, etc.

–  Sales Reporting

⦿ Set bespoke rules for various data processing engines, from data absorption through cleansing, transformation and loading on Snowflake

Outcomes

>70% Accuracy of insights

~60% Reduction in operation costs

3600 view of customers, enabling personalized offerings and responses

TECH STACK
AWS S3AWS DMSAWS Glue
AWS Secrets ManagerAWS RDSAWS IAM
AWS SESCloudWatchSnowflake
Tableau

Leave a comment

error: Content is protected !!
Explore
Drag