TapLive

Problem Statement

DescriptionTapLive is a decentralized, location-based livestream dispatch platform that connects people with real-world actions in real time. The platform allows users anywhere in the world to request on-demand livestreams from specific physical locations. Local users on-site can accept the order, start streaming, and deliver real-time visual information to the requester.TapLive’s core goal is to build a trustable human network layer on top of real-world geography. Instead of passively watching public streams, users can initiate their own “live missions” — for example:See a famous landmark live from another countryCheck real-time local conditions before travelingVerify real-world situations in person (e.g., line length, weather, events)Our MVP focuses on three key components:LBS-based task matching — orders are matched to nearby available streamers in real time.Livestream module — powered by low-latency video technology to deliver stable and interactive streams.Order system — requesters can issue orders and streamers can accept them instantly, enabling peer-to-peer collaboration.We prioritize user trust, compliance, and risk control from day one. Every interaction is verifiable, and the system encourages transparent and accountable collaboration between strangers.Future phases will gradually expand the ecosystem — but our hackathon build focuses on delivering a simple, working MVP demonstrating how location-triggered live orders can bridge the digital and physical world.

Solution

We built TapLive MVP as a lightweight, location-based livestream dispatch platform designed to work reliably under real-time conditions.Our tech stack is centered on a React + Node.js + Express architecture, using TypeScript end-to-end for maintainability. The frontend runs on a Vite build system to achieve fast dev cycles, while the backend handles real-time order matching, basic authentication, and session control.For live video, we integrated LiveKit to enable low-latency, bidirectional streaming with interactive features (chat, status updates). LiveKit’s WebRTC foundation allows us to scale livestream sessions without depending on centralized infrastructure.Core components:Frontend: React (Vite) + Tailwind CSS for UIBackend: Node.js / Express with REST APIsVideo: LiveKit WebRTC for livestreamingDatabase: PostgreSQL for persistent order/session dataGeo layer: browser-based Geolocation API + LBS task allocation logicWe implemented a minimal LBS dispatcher that pairs incoming orders with streamers in the same or nearby geographic area. A real-time event bus broadcasts order status changes and enables peer-to-peer coordination.One of the more “hacky” but effective things we did was to simplify geolocation matching into grid zones — reducing the overhead of calculating exact distances while still keeping the system responsive. This allows us to simulate large-scale order matching during the hackathon without requiring a full GIS stack.Partner technologies like LiveKit drastically reduced the time required to build a stable streaming experience, letting us focus on task orchestration and risk control logic rather than low-level streaming infrastructure.