Time
2 weeks
Location
Malta
Sector
FinTech
Description
The founder of a financial services company faced the crucial challenge of automating the manual validation of sales agent conversations. The objective was to detect forbidden phrases and potentially risky business communication patterns, a critical requirement due to regulatory restrictions in the financial services sector.
Solution
To address this regulatory challenge, we devised an automated solution that harnessed the power of machine learning. This solution focused on the analysis of call transcripts, the identification of high-risk patterns and key phrases, and worked with large-scale machine learning models trained on previously manually labeled data. Additionally, we created a user-friendly interface to facilitate the easy review of transcripts and the identification of potentially dangerous patterns.
Results
The implementation of this system delivered a remarkable reduction in the manual monitoring of communications, slashing the effort required from about 200 hours to merely three hours per month. This substantial reduction in manual labor not only led to significant cost savings but also enabled the company to scale its operations more efficiently, while ensuring compliance with regulatory requirements.