Every time you tap through a subway gate in New York City, you’re using Kotlin Multiplatform! In this episode, we sit down with Paweł Kwieciński from Masabi to explore how they brought a decade-old Java codebase to KMP, powering ticket validation across mobile apps, embedded devices, and backend systems for public transit agencies worldwide.

Resources:

• Masabi
• J2ObjC (Java to Objective-C translator)
• Google Workspace Migration to Kotlin
• Large Scale Changes with AI – Uber | Ty Smith
• KotlinX DateTime
• Ktor (Client & Server)
• Burst library (archived)
• Java ME (Micro Edition)
• Mbed TLS
• SQLDelight
• Compose Multiplatform
• JetBrains AI Assistant & Junie

Hosts:

• Sebastian Aigner – Website | Bluesky
• Márton Braun – Website | Bluesky

Guest:

• Paweł Kwieciński

Timeline:
(0:00) Introduction
(1:12) What is Masabi?
(3:03) How does Kotlin fit into this?
(6:30) System architecture overview
(8:33) Embedded SDK and mobile apps
(10:33) Ticket inspectors and validators
(13:22) Offline operation
(16:29) The KMP decision
(24:42) Evaluating alternatives
(27:21) Planning the migration
(33:28) First lines of Kotlin
(43:45) Translating Java to Kotlin
(53:22) AI-assisted code translation
(56:39) Early AI era challenges
(1:05:01) Testing challenges
(1:08:22) KMP dependencies
(1:11:34) Domain-specific dependencies
(1:16:55) Java ME and C interop
(1:23:22) Backend-client sharing
(1:26:59) Current status
(1:31:48) Future plans: CMP
(1:36:26) Wrap-up