Sumitomo Mitsui Bank has turned to Silicon Valley data wrangling company Trifacta to slash preparation times for the aggregation and analysis of mushrooming customer data troves.
The Japanese bank is planning to roll out Trifacta's technology across the board after running field tests late last year which demonstrated that it would be possible to cut the work times for data preparation by around 80%.
SMBC already employs AI to scan and pick up patterns in customer data, but relies on manual data entry for the collation and preparation of relevant data sets to feed into the machine.
"At SMBC we have been employing AI to raise the efficiency and sophistication of data/analysis processing, but this procedure is being performed manually and involves trial and error," states the bank. "It takes time, so we have been relying on the knowledge and experience of experts to complete it."
SMBC says that procedures for converting millions of lines of customer data into a format that can be analysed is responsible for 30% of the workload of its entire data-analysis operation.
Trifacta works across multiple databases, capturing and reformatting relevant data without the need for specialist skills.
The deployment, which will be supported by the Japan Research Institute and NEC, marks Trifacta's first customer engagement in Japan.
Trifacta develops data wrangling software for data exploration and self-service data preparation for analysis.