P2P Decentralised Messaging Platform Development

Job ID: 40595449

Budget: ₹1,500 – ₹12,500 INR

I’m ready to take the classic WhatsApp-style messenger a giant step further by launching a fully decentralised, peer-to-peer chat platform that runs smoothly in any modern browser and installs as a progressive Web App. The entire interface must blend neomorphism with glassmorphism, delivering a slick, tactile look while remaining lightweight and accessible on both desktop and mobile screens.

Core requirements
• Architecture: true P2P—no central server for message storage. End-to-end encryption, WebRTC (or an equivalent protocol) for real-time transport, and local data persistence that effortlessly syncs when peers come back online.
• Platforms: Web + installable PWA.
• Sign-in: email/password plus optional device biometrics (Face ID, Touch ID, Windows Hello).
• Messaging features: text, voice snippets, and high-quality video calls handled inside the same conversation flow.
• Built-in AI: each client keeps a lightweight model that fine-tunes locally, with encrypted parameter sharing so improvements propagate through the network without exposing private data.
• UI/UX: complete component library in the hybrid neo/glass style, dark & light themes, accessible colour contrast, smooth micro-interactions, and responsive states for all breakpoints.

Deliverables
1. Source-controlled codebase (frontend, P2P layer, AI module, cryptography utilities).
2. Design assets: Figma (or similar) file plus exported icons/illustrations.
3. Setup scripts and a brief deployment guide so I can spin up a local node and invite testers instantly.
4. A short video walkthrough proving text, voice, and video all function under a NATed network scenario.

Acceptance criteria
– New user can register with email, then opt-in to biometric sign-in.
– Two fresh browsers on separate networks exchange keys and start an encrypted video call in under 5 seconds.
– Killing the connection for one peer queues messages and re-syncs automatically on reconnection.
– Switching between light/dark themes never exceeds 16 ms frame budget.
– AI replies appear in-context and adapt after three rounds of feedback without cloud interaction.

If you’re confident working with WebRTC, service workers, IndexedDB (or similar), and modern CSS/Three.js-powered visuals, let’s make this next-gen messenger a reality.