Posted by Thomas Ezan, Senior Developer Relations Engineer and Tracy Agyemang, Product Marketing Manager
carrot is a hyperlocal, community-driven peer-to-peer marketplace app that enables users to buy, sell, and trade items with other verified users. Since launching in South Korea in 2015, the platform has expanded into global markets, with over 43 million registered users.
After launching in North America, Carrot’s engineers noticed that 30% of users in the region used a non-English device language, such as Spanish. To make the app more accessible, the team wanted to bring instant and largely seamless translation functionality to Carrot. The developers determined that the most effective way to implement quality translations would be to integrate the AI ​​service directly into the app, so they selected Firebase AI Logic and its Android SDK to access Gemini Flash Lite, which led to higher purchase conversions among non-English users.
Integrating Gemini Firebase AI Logic
The team initially tested two on-device options: ML Kit Translation SDK and Gemini Nano. But the team found challenges with each: ML kit translation did not meet the team’s quality expectations, and Gemini Nano required the user to download model data if it was not already on the device.
The team then tested the Firebase AI logic. By calling the Gemini API directly from the app, Firebase AI Logic delivered accuracy at a speed that mirrors a natural conversational cadence.
According to Taegyu An, Android software engineer on Carrot’s mobile platform team, integrating Firebase AI logic into the app was a “remarkably straightforward experience.” TaeGyu and team used the platform documentation And code samples to create a proof of concept in less than three hours.
This allowed the team to spend more time refining the signals and finding optimal configuration values. “Even without extensive experience in writing official documents, Guides and Tips “It’s easier to quickly identify the right direction to improve translation quality,” said Wonjeong Lee, Android software engineer on Carrot’s North America product team.
This low barrier to entry and fast turnaround time enabled engineers to keep development costs low and go from proof of concept to production code in just two weeks – all without setting up a dedicated backend. This also freed up time to focus on UX and policy design, such as opt-in behavior and translation banner terms.
Increase sales with advanced AI features
Since implementing translation using Gemini and Firebase AI Logic, the Carrot team saw higher purchase conversions among non-English users, indicating that the translation feature is helping increase sales.
Among users who used a non-English device language, One in three of those who were shown the translation banner actively used the feature. The team also noticed that buyers were offering translation functionality 2.4 times more likely to initiate a chat with a vendor compared to those who were not.
The flexibility and simplicity of deploying Firebase AI Logic has inspired the team to explore other features to simplify the workflow of their engineers. TaeGyu concluded, “It is beneficial to build features across various Android devices at scale while helping neighbors connect and interact within their local communities.”
Going forward, the team is planning implementation Server Prompt Template To adjust prompts after release without shipping a new version of the app. This, combined with remote configThis should help the team iterate faster and reduce operational overhead.
get started
Learn how to build Gemini-enabled features like AI translation and in-app personalization and more firebase ai logic To provide our users with a better experience faster.
Carrot Marketplace, Hyperlocal P2P Apps, Gemini Flash Lite Integration, Firebase AI Logic for Android, AI-Powered App Translation, Global Expansion of South Korea Marketplace, Peer-to-Peer Trading Accessibility, Machine Learning for Mobile Apps, Boosting Purchase Conversions with AI, Gemini API for Developers, Firebase SDK Translation, Server Prompt Templates, Remote Config Android, Non-English user engagement, Gemini Nano vs. ML Kit, Carrot North America expansion, community-driven marketplace technology.
