GitHub - wesleyzecaamino-rgb/project-baobab: Decentralized Edge AI Mesh Orchestrator with Automated PayPal Micro-Payments

Why it matters

Project Baobab proposes an alternative to centralized AI compute by sharding inference workloads across a peer-to-peer mesh of low-power edge devices, including Raspberry Pi 5 single-board computers. Where conventional AI infrastructure depends on energy-intensive data centers with heavy grid draw and high operating costs, this orchestrator distributes tasks across laptops, mini-PCs, and off-grid hardware, some powered by solar panels and battery storage. The system's defining feature is its payment layer: when a node completes a verified task, the orchestrator authenticates with PayPal using an OAuth client-credentials flow and creates a checkout order through the payments provider's REST interface, issuing an instant micro-payout. Written in Python and using the language's thread-based concurrency tools, the design dispatches prompts to multiple nodes simultaneously, which its author says removes single points of failure and supports modular expansion. The cross-platform code runs on Linux distributions including Raspberry Pi OS. For Veriwire readers, the story matters as a test of whether hobbyist-scale hardware and automated settlement can together sustain a genuinely decentralized, self-funding AI infrastructure model.