Samiksha Verma@samikshaverma

Problem: Seniors in India are prime targets for digital fraud — fake KYC alerts, OTP scams, courier scams, and fake authority calls. Limited tech literacy and English-only tools leave them unable to tell a real alert from a scam, and by the time family finds out, the damage is often done. Solution: Digital Sathi SafeSathi is a bilingual (Hindi/English/Hinglish), voice-first AI safety companion built for senior citizens and their families. Users can type a suspicious message, speak it aloud, upload a screenshot, or describe a phone call — SafeSathi analyzes it using a trained ML model (TF-IDF + classifier) combined with rule-based checks for urgency language, link risk, and known scam patterns (Bank/KYC, OTP, courier, fake police, lottery, tech support). Within seconds, the user gets a clear verdict — Safe / Suspicious / Likely Scam — along with the scam probability, scam category, and, most importantly, plain-language reasons why it was flagged (not just a black-box score). Each result includes concrete safety tips and a "Read Aloud" option in Hindi or English, so literacy and vision are never a barrier. A "Share with Family" button lets users forward the full result instantly via WhatsApp or any app, so family members can step in immediately if something looks wrong — turning an individual moment of doubt into shared protection. Beyond scam checking, SafeSathi includes a Learn section explaining how common scams work, and Digital Sathi Guides — step-by-step, bilingual walkthroughs for everyday digital tasks seniors often fear (sending money via UPI, paying bills, video calling on WhatsApp, booking IRCTC tickets) — each with a built-in safety reminder (e.g., never share your UPI PIN). An AI chat assistant is available on every page for open-ended questions in Hindi or English, and even resolves "play a song" requests directly to YouTube — all while being explicitly scoped to never ask for OTPs, PINs, or banking details. Target users: Primarily senior citizens navigating digital payments and messaging; secondarily their families, who gain a shared tool to verify threats and support their parents/grandparents remotely. What makes it unique: Most scam-detection tools are English-only, text-input-only, and give vague "risky" alerts with no explanation. SafeSathi is built senior-first: multiple input modes (text, voice, screenshot, call description), full Hindi/Hinglish support, transparent AI reasoning, voice output, and a family-sharing loop — turning digital safety into a shared, understandable experience rather than a lonely guessing game.

💻 Online Hackathon | ⏳ Build: 48 hours | 📅 September 25–27, 2026 The official IBM Bob 2.0 hackathon. Build what's next in AI-assisted development, alongside the developers and experts shaping it. ⭐ Get help from experts throughout the event 🤝 Join a team or build solo 💰 Prize pool: $12,000 🧑🏻‍💻 Register before the Kick-Off Stream to start building with Bob