The ubiquitous soundbox, a small device that loudly confirms Unified Payments Interface (UPI) transactions, is poised for a significant upgrade in India. Beyond simply announcing successful payments, these AI-powered devices are being developed to function as mini bank branches, offering a suite of financial services directly to merchants.
From Payment Confirmation to Banking Assistant
For millions of Indian merchants, from street vendors to high-street retailers, the soundbox is an indispensable part of daily commerce. However, when a payment fails or a settlement is delayed, the current devices offer little immediate support, often requiring merchants to consult an app or call a helpline.
Bengaluru-based ToneTag is at the forefront of this transformation. In July 2026, the company unveiled eKosha, a voice-first AI banking assistant specifically designed for merchants. Building on its decade-old sound-wave protocol, which converts data into sound for device communication, eKosha processes voice commands locally, enabling merchants to inquire about settlements, request cheque books, or seek loan information without an app or bank visit.
Expanding the Digital Frontier
ToneTag isn't alone in this venture. In October 2025, Paytm launched its own AI soundbox, capable of operating in 12 languages and answering questions about collections, trends, and business performance. Paytm's founder, Vijay Shekhar Sharma, hailed it as a "smart and capable business partner" for small businesses.
The broader push towards voice and AI is seen as the next major step for UPI. Dilip Asbe, MD and CEO of the National Payments Corporation of India (NPCI), suggests AI could bring the next half-billion users into the digital payments ecosystem and play a vital role in fraud detection, credit distribution, and multilingual interfaces.
Addressing Critical Security Challenges
While the potential for voice-first banking is immense, it also introduces new security concerns. Amit Jaju, Senior Managing Director at Ankura Consulting, highlights three key risks:
- Ambient noise attacks: Commands could be mimicked or replayed in a noisy environment.
- Social engineering: Merchants might be tricked into repeating phrases that are later used fraudulently.
- Connectivity compromise: A soundbox's network could be hijacked to inject spoofed audio.
Jaju emphasizes that "voice interfaces confirm intent, not identity." To mitigate these risks, he proposes a three-layered protection system: device-bound authentication, mandatory secondary confirmation for money-moving transactions (like a PIN or app approval), and behavioral monitoring to flag unusual patterns.
ToneTag's solution embeds the merchant's identity within their voice itself. During registration, a merchant's phone number, device, and voice are linked, meaning their unique voice signature travels with every command, eliminating the need for separate codes.
The Future of Merchant Banking
For banks, the adoption of these conversational devices offers significant scalability and cost-efficiency. While many banks already integrate soundboxes into their merchant offerings, the services often remain confined to apps. The crucial question for companies like ToneTag is whether these voice-first interfaces can transform the soundbox into a truly standalone and meaningful banking channel, rather than just a smarter payment terminal, for India's 65-80 million MSMEs.