From Midnight Power Cuts to Multi-Agent Voice AI: How I Built Raksha in 10 Days
Building voice AI sounds deceptively clean on paper: capture speech, stream it to an STT engine, prompt an LLM, and synthesize audio back in real time. In reality, building a real-time, multilingual voice assistant that handles high-stress banking fraud while running on desktop hardware in 30°C heat with unstable power is a completely different story. As a Class 12 student diving into my…
India's digital payments sector is expanding rapidly, but so too are instances of financial fraud. Scammers inundate citizens with urgent SMS messages demanding immediate action, often in the form of OTPs (one-time passwords). During times of panic, people are more likely to follow verbal instructions rather than read lengthy text advisories. To address this issue, a Class 12 student built a voice assistant called Raksha for the #VoiceForBharat Challenge.
Raksha is designed to protect Indian citizens from cyber scams, verify official government schemes, and notify human coordinators about active financial fraud. The student faced numerous challenges during the 10-day development process, including power outages, technical issues with the speech-to-text (STT) engine, and the need for strict privacy and security measures. Despite the obstacles, Raksha was able to emerge as a multi-agent voice assistant, capable of handling high-stress situations in real-time.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
