BodegAI is on a mission to give independent convenience stores, bodegas, delis, and liquor stores the same data tools that large retail chains have. FlyDevs has been their engineering partner since day one — building the full Kotlin POS system from the ground up.
FlyDevs joined as BodegAI was pivoting toward an AI-powered POS. First decision: Kotlin on Android as the primary platform — mature ecosystem, strong hardware support for POS peripherals, and the right tradeoff between flexibility and deployment speed.
Built the core point-of-sale flow: smart product scanning with automatic SKU/price/product recognition, pre-loaded catalog of 200K+ products ready from day one, and dual pricing for automatic cash vs. card price management.
Real-time inventory management with stock alerts and cloud sync. CRM layer for customer behavior tracking and relationship tools. Remote management dashboard giving store owners full visibility from anywhere, 24/7.
Automated sales analysis and business optimization built into the AI layer. Real-time reporting dashboards for product performance and business insights. FlyDevs continues shipping as BodegAI scales across the US.
POS hardware (barcode scanners, receipt printers, payment terminals) requires tight native integration. Cross-platform frameworks all had driver compatibility gaps with the hardware BodegAI needed to support. Native Kotlin gave us full control over peripheral I/O with no abstraction layer in the way.
Most POS systems require store owners to manually enter every product. We built a pre-populated catalog covering the SKU universe relevant to bodegas and convenience stores — so stores can be operational from day one without a data entry burden.
We built the POS foundation to be rock-solid first — accurate transactions, reliable inventory sync, clean data model. AI features (sales pattern detection, reorder suggestions, anomaly flagging) were layered on top once the underlying data was trustworthy. AI built on dirty data is a liability.
When customers paid part cash and part card, the dual pricing logic had to split the discount correctly across line items. Early versions had rounding errors that caused cent-level discrepancies — harmless in isolation but problematic at audit time. We rewrote the pricing engine with explicit split-tender test cases covering every combination before it touched a production terminal.
200K+ SKUs
2 industry wins
Kotlin (Android)
Ongoing
Book a free 20-minute call. We'll scope your project, recommend the right team composition, and give you a realistic timeline — no pitch, no pressure.
Book a Free Call →