Location :
Mountain View, CA (On-site preferred, at least 4 days / week)
Position Summary :
We are looking for a highly capable engineer to design and deploy real-time, on-device machine learning solutions for mobile devices. This role focuses on developing privacy-first, resource-optimized ML systems that operate directly on Android hardware, supporting high-impact AI applications in real-world environments.
The ideal candidate brings deep technical expertise in on-device intelligence and mobile ML pipelines, and thrives in fast-paced environments that demand performance, security, and adaptability.
Key Responsibilities :
- Design and deploy efficient on-device ML models tailored for Android platforms.
- Build end-to-end ML pipelines using :
TensorFlow Lite
ML Kit (including GenAI APIs)MediaPipePyTorch MobileOptimize models using techniques like quantization, pruning, and distillation to meet mobile performance targets.Develop context-aware, real-time inference systems and data pipelines on-device.Implement privacy-first architecture, ensuring all data processing and inference is local.Collaborate with backend / cloud teams to integrate model orchestration systems (e.g., MCP, Vertex AI, SageMaker) for :Model delivery and remote updates
Telemetry and performance monitoringA / B testing and rollout strategiesImplement secure storage and encrypted data handling in line with privacy and compliance standards.Support adaptive model behavior using on-device personalization, federated learning, or similar privacy-preserving techniques.Technical Requirements :
Must-Have Skills :
Strong Android development experience using Kotlin and / or JavaProficiency with on-device ML tools :TensorFlow Lite
ML KitMediaPipePyTorch MobileSolid understanding of mobile constraints :Real-time inference
Low-latency processingModel size and resource optimizationExperience in integrating mobile apps with backend / cloud systems for :Model lifecycle management
Secure telemetry and data analyticsKnowledge of Android security best practices, including sandboxing, permissions, encryption, and local data protectionNice-to-Have Skills :
Experience with federated learning, differential privacy, or on-device personalizationFamiliarity with cloud infrastructure (e.g., AWS, GCP) and ML deployment workflowsBackground in mobile AI features like anomaly detection, behavioral modeling, or privacy-focused applicationsExperience with model orchestration platforms such as MCP, Vertex AI, or SageMakerEducation & Experience :
Master's degree with 5-7 years of relevant experience, orPhD with 3 years of relevant experiencePreferred : Less than 10 years of total professional experienceWork Schedule :
On-site presence preferred - at least 4 days / week in Mountain View, CAStandard business hours, minimal overtime except during key sprintsInterview Process :
1 Technical Phone Screen2 Virtual Technical InterviewsPosition Type & Growth Potential :
Contract role with high potential for extensionStrong possibility of full-time conversion based on performance and business needsCore Technical Keywords (for resume alignment) :
TensorFlow Lite (TFLite)ML KitMediaPipePyTorch MobileOn-device machine learningMobile ML pipelineEdge AIModel quantization / pruning / distillationReal-time inferenceFederated learningDifferential privacyTelemetry integrationSecure Android developmentModel orchestrationCloud-integrated ML (e.g., Vertex AI, SageMaker, MCP)