Best Ring POS - Offline-First Festival Payments
Completed: 2017
Architected offline-first payment sync for festival POS, and ran it live at Lollapalooza-scale events — transactions kept processing through outages, with fraud data synced down opportunistically to disable compromised cards before they cleared.
The Problem
Festival point-of-sale has to keep processing payments through internet outages at scale — a dead connection at peak volume isn't a degraded experience, it's a lost sale and a security gap, since compromised cards need to be caught and disabled in real time even while offline.
The Solution
Architected offline-first payment sync for festival POS: terminals cached and batched transactions locally while offline, and as connectivity became available (polling opportunistically — a dedicated mesh network wasn't actually available, though the design would have worked over one too) synced down updated fraud-detection data so newly reported stolen/compromised cards stopped clearing at any terminal, not just the one that caught the report. Acted as CTO / Chief Software Engineer for this five-month engagement (2017) — building the team, running the SDLC, hiring interns and FTEs, drafting PCI-compliance architecture changes, and running the system live at large-scale festivals including Lollapalooza. This was pre-AI-tooling, and every payment/reporting provider wanted a differently-shaped report, so a meaningful part of the job was scripting and generating real-time metrics dashboards on the fly during events, then exporting/saving custom reports for each provider afterward.
The Impact
Kept payment processing running through connectivity outages at large-scale festival events including Lollapalooza, with synced fraud data reducing damage from stolen/reported cards festival-wide rather than terminal-by-terminal. No published dollar-figure metrics on file for this 2017 engagement.