TD Bank: building the bank of the future with behavioral analytics and sentiment analysis
Key highlights
Category
Banking/Financial Services
Location
Headquarters: Toronto, Ontario, Canada
Solution highlights
- Modern Data Platform: Cloudera Enterprise
- Workloads: Data Science & Engineering, Analytics Database
- Key Components: Apache Hive, Apache Impala, Apache Spark, Cloudera Manager, Cloudera Navigator
- BI and Analytic Tools: AtScale, Tableau
- Data Science Tools: Scala, Python
Applications supported
- Business Intelligence
- Propensity analysis
- Sentiment analysis
- Sequence analysis
- Fraud detection
Impact
- Reduced repeat customer complaints by 30 percent
- 90 percent productivity improvement for analytics projects
- Reduced data management costs by 60 percent
- Decreased storage costs by 98 percent per TB
TD Bank delivers an exceptional customer experience and convenience across channels with deeper visibility into each customer’s journey.
Headquartered in Toronto, Canada, TD Bank Group offers a full range of financial products and services to more than 25 million customers worldwide.
Challenge
TD Bank wanted to integrate data from across its business units and use advanced analytics and machine learning for a variety of activities--from marketing to fraud detection.
“Our mission is to provide legendary customer experience to all of our customers,” said Joe DosSantos, vice president, Enterprise Information Management, Technology Services at TD Bank. “Data and analytics are really at the heart of becoming the bank of the future.”
Solution
TD is implementing a modern data platform for analytics and machine learning that ingests and analyzes data from over 100 corporate systems for use in customer marketing, fraud analytics and cybersecurity. By moving from a traditional siloed environment to a modern architecture on Cloudera, the organization reduced data management costs by 60 percent and storage costs by 98 percent per terabyte (TB).
“Cloudera allowed us to bring thousands of data sources together in one place at a fraction of the cost,” said DosSantos. “That allows us to get the data into the hands of those who need it much more quickly, and much more cost-effectively.
Results
“Understanding your customers begins with understanding sentiment analysis, behavioral patterns, and predictive modeling,” said DosSantos. “We've done extensive work to make sure that we understand what our customers are looking for and where they are on their respective journeys.”