Machine Learning Engineer, hibrido

Our first application is an agent-first mobile communication app. We believe building applications that users use daily for communication, coordination and collaboration needs will help integrate AI into their lives. The net effect is to eliminate misinformation, achieve low coordination costs, enable high quality tasks and collaboration at scale with minimal or little prompting. Our first application is an agent-first mobile communication app. We believe building applications that users use daily for communication, coordination and collaboration needs will help integrate AI into their lives. The net effect is to eliminate misinformation, achieve low coordination costs, enable high quality tasks and collaboration at scale with minimal or little prompting. Work on classification, retrieval, ranking, personalization and evaluation, taking ownership from model development to production deployment. This role is for engineers who enjoy solving practical ML problems, improving model quality and shipping reliable systems used by real users. Improve search relevance and context retrieval using embeddings, hybrid search and reranking. Develop evaluation pipelines, labeled datasets and metrics to measure model quality. Fine-tune and adapt existing models when supported by sufficient data. Build and maintain ML training, inference and deployment pipelines. Debug production issues, monitor model performance and ship improvements based on real-world feedback. Python PyTorch / JAX Production ML systems and inference pipelines Strong foundations in machine learning and modern neural architectures. Hands-on experience building and deploying ML models in production. Understanding of model evaluation, experimentation and performance metrics. Comfortable writing production-quality code and working with real-world datasets. Able to balance model quality, latency, cost and reliability. ML models in production meet expected accuracy, relevance, latency and reliability targets. Classification and ranking improvements deliver measurable gains in model quality. Training, inference and data pipelines are robust, reproducible and maintainable. Collaborates effectively with engineering and product teams to deliver reliable ML-powered features. We make decisions collectively and move quickly, balancing high-quality work with continuous learning.