Sagar Yadav@sagargyadav874

The fine print is where websites get you: a loan advertised at "1.5% a month", a protection plan ticked before you arrive, a fee that only shows up on the last step, a cancellation clause buried in section 7. Iris is a voice copilot that reads the page with you, so you notice these things before they cost you. Iris lives in Chrome's side panel. You talk to it the way you'd ask a friend looking over your shoulder. "What does this loan really cost me?" "Can I cancel anytime?" "This looks phishy, can you check?" It answers from the page itself, quotes the exact clause and highlights it. It also gives you a heads-up at four moments: Before you fill: it guides you through confusing forms, part by part, and explains a field when you ask. Before you agree: it answers questions about the terms from the actual clause. Rest the mouse on a clause and tap Explain. Before you pay: it speaks up on its own about pre-ticked add-ons, free trials that turn into subscriptions and late fees, and works out the true yearly cost of flat-rate loans. Before you trust: it checks how old a site's domain is, whether its name looks like a known brand's, and whether there are public scam reports. Ask for details, and it looks the fraud up on the web. How it's built: AssemblyAI's Voice Agent API handles listening, turn-taking, interruptions and Iris's voice, and calls Iris's tools as client-side tools. Claude, through AssemblyAI's LLM Gateway, reads long terms (Sonnet 5) and answers quickly (Haiku 4.5). Tavily searches the web for error codes, terms and scam reports. RDAP gives a domain's age. Costs, dates and domain ages are always computed in code, never by a language model. Privacy is designed in: Iris never clicks, types or submits. It only knows whether a form field is filled, never what you typed. Page text is kept for 30 minutes in memory and never logged. The demo uses QuickCred, a fictional lender with realistic traps, plus real websites. Full Demo - https://youtu.be/wzjqNlHt7aI

💻 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

🌎 Online Hackathon | 💻 Month-long challenge | Dates: Sep 1–30, 2026 | 🏆 $10,000 Prize Pool ($5k cash + $5k in AAI credits) A month-long online challenge run by lablab.ai together with AssemblyAI, the Voice AI infrastructure company for builders. Every participant builds on AssemblyAI. You can join at any point — registration stays open for the whole build window, so you can start on day one or pick the project up halfway through and still submit. 🌍 Fully online hackathon. Join and build from anywhere in the world.