Android Software Engineer (AI)

Location: Remote from China; address in Beijing, China. Interviews may be conducted virtually or onsite.

Company

Building proactive AI-native applications that help users manage conversations, errands, organization, and workflows with minimal prompting.

What you will do

• Build and maintain production Android applications using Kotlin.
• Integrate AI-powered chat, vision, voice, and recommendation features through backend APIs.
• Design UX patterns for streaming responses, retries, partial results, and other AI interactions.
• Optimize performance, memory usage, CPU efficiency, and responsiveness for AI-heavy flows.
• Implement analytics, logging, and feedback capture to support AI evaluation and iteration.
• Collaborate with backend and ML engineers on API contracts, system behavior, stability, security, and scalability.

Requirements

• 3+ years of Android development experience with Kotlin.
• Hands-on experience integrating AI features such as LLM, vision, or speech APIs.
• Strong understanding of asynchronous programming with Coroutines and Flow.
• Experience with REST or gRPC APIs, structured data formats, and SQL or NoSQL databases.
• Strong debugging and performance profiling skills, including work with latency, partial failures, and non-deterministic behavior.
• Published production applications on the Google Play Store; experience with MLKit or light on-device inference.

Nice to have

• Experience with TensorFlow Lite and on-device inference.
• Experience with Kotlin and Java in production Android applications.

Culture & Benefits

• Work with a small, high-talent-density, hands-on team.
• Collective decision-making combined with rapid execution and independent judgment.
• Focus on balancing high-quality delivery with continuous learning.
• Prompt hiring decisions and a streamlined interview process.

Hiring process

• Applications are reviewed by technical team members.
• Expect three interviews, with no more than four if there is a strong fit.
• Interviews take place via virtual meetings and/or onsite, followed by a prompt decision.