Mitigating cross-border data risks with secure on-premise infrastructure and local language processing

Every day, thousands of customers in Bangladesh contact banks, financial institutions, e-commerce, transport, and healthcare providers through call centers, sharing sensitive details like phone numbers, dates of birth, transaction history, and voice samples. Although Bangladesh Bank mandates storing financial customer data on local servers, many institutions still rely on foreign cloud services, risking cross-border data leakage. Addressing this infrastructure and security gap, homegrown software company VivaSoft Limited has introduced 'Kothon', a universal contact center platform that integrates call management, conversation analysis, and data protection into a single system.

A key feature of Kothon is its flexibility to be deployed within an organization’s internal infrastructure, such as on-premise hardware, private clouds, or local data centers. This localized deployment minimizes reliance on foreign or public clouds, giving institutions absolute control over where their data resides, who accesses it, and how it is utilized. For robust data protection, this AI-driven suite includes advanced features like encryption, audit logs, permission-based access, and sensitive data masking, which protects financial details like account numbers in written records.

Beyond security, Kothon significantly enhances operational efficiency. In traditional call centers, verifying a customer's identity can take 1 to 2 minutes, which consumes a major portion of a standard five-minute call. Kothon slashes this wait time by automatically recognizing registered customers through their mobile numbers. The system can handle thousands of concurrent calls and automatically resolve repetitive queries. While recruiting and training 50 human agents can take 3 to 6 weeks, Kothon seamlessly manages holiday rushes without the need for additional manpower, though human agents remain integral for complex issues.

Kothon also transforms call centers into valuable sources of business intelligence. For instance, Autonemo, a local fleet management company, used Kothon to analyze call logs and discovered that nearly 20 percent of callers—mostly small-scale local entrepreneurs owning a few trucks—were repeatedly asking for a single-account multi-truck management feature. While human agents failed to recognize this underlying pattern buried in thousands of calls, Kothon's analytics tools successfully extracted this crucial business insight.

Engineered specifically for the Bangladeshi linguistic context, Kothon can seamlessly comprehend Code-Mixed language, where Bengali and English are blended (e.g., "আমার লাস্ট ট্রানজেকশনটা একটু চেক করে দেন"). Furthermore, the platform filters out background noise from busy environments and accurately interprets various regional Bangladeshi accents. Shafiul Hasan, Co-founder and CEO of VivaSoft, emphasizes that call centers should be viewed as strategic decision-making assets rather than mere cost centers. Technologies like Kothon indicate that the future of customer service lies at the intersection of data security, local language adaptation, and human-tech collaboration.

//DBTech/PAS/MI//