Android Application Engineer

LILT is looking for an Android Application Engineer to join the team building LILT Converse, an on-device instant translation application. Converse enables secure, real-time speech translation without an internet connection, designed for government, defense, and enterprise teams operating in communication-sensitive or connectivity-constrained environments. The role focuses on building software that runs large multilingual speech models directly on Android hardware leveraging Qualcomm Snapdragon AI acceleration.

Description

As the Android engineer on Converse, you will own the Android application end-to-end — from architecture and on-device ML integration through performance tuning, release, and reliability in the field — working closely with ML, product, and design on a small senior team. You will architect new features, integrate on-device ML models, optimize performance, manage large local model assets, and ensure reliability in production environments.

Responsibilities include delivering multilateral translation (3+ simultaneous languages), utterance management, kiosk deployment modes, enterprise configuration, reasoning about hardware constraints, APK packaging, sideloading workflows, and managing multi-gigabyte inference workloads on mobile SoCs.

Requirements

Required

• 4+ years of professional Android development (Kotlin) with end-to-end ownership of a production app
• Strong Android fundamentals: Android SDK, Jetpack, lifecycle, background work, storage APIs
• Experience integrating native libraries via NDK/JNI
• Performance optimization on physical devices (memory, latency, thermal behavior)
• Experience with offline-first applications and large on-device assets
• Real-time audio implementation on Android: microphone capture, TTS/speaker output, low-latency audio pipelines, USB microphone support
• On-device speech pipeline integration (ASR → translation → TTS), streaming partial results, live conversation UI turn-taking
• Enterprise deployment familiarity: sideloading, MDM/kiosk modes, pre-provisioned device configuration
• U.S. citizenship or eligibility for required government access authorization

Strong Plus

• On-device ML runtime integration (ONNX Runtime, TensorFlow Lite, Qualcomm QNN / Snapdragon NPU acceleration)
• External display / kiosk UX (HDMI output, overlays)
• Secure data handling: encrypted export, USB storage workflows, air-gapped environments
• Experience with government, defense, or enterprise customers

Bonus

• Deep speech/audio domain expertise (latency/quality tuning, noise handling, audio preprocessing, model quantization on mobile SoCs, speech model runtimes)
• iOS / Swift experience
• CI/CD with hardware-in-the-loop testing on physical devices

Compensation

Compensation Range: $143K - $206K

• Tier 1: $157K – $206K
• Tier 2: $143K – $192K