Machine Learning Engineer Location: Remote (LatAm time zones preferred) Type: Full-time Reports to: CTO About Mappa Mappa builds real-time voice-AI infrastructure that decodes over 100,000 behavior patterns per person—based on prosodic and linguistic signals—and turns them into actionable insights for hiring and beyond. Our ML stack is core to transforming raw audio into structured behavioral data at scale. Role Purpose Design, train, and deploy machine learning models that power Mappa’s behavioral inference engine. You'll work across the ML lifecycle—from data ingestion to optimized neural network deployment—ensuring speed, precision, and reliability at scale. Key Outcomes (First 6 Months) Develop and productionize at least two core models contributing directly to behavior extraction. Optimize inference performance by reducing model latency and memory consumption by ≥ 30%. Implement a modular training pipeline that supports continuous experimentation and deployment. Collaborate with data and infra teams to streamline the ML-to-production workflow using MLOps best practices. Core Responsibilities Design, train, and evaluate ML models—especially neural networks—for behavioral and linguistic data. Optimize tensor operations for efficient computation and memory usage in real-time environments. Implement scalable pipelines for data preprocessing and feature engineering using tools like Spark or Kafka. Apply model evaluation techniques, error analysis, and performance monitoring in production. Collaborate with Engineering and Data Science to integrate models into the voice-AI infrastructure. Contribute to MLOps processes including CI/CD for models, versioning, and experiment tracking. Must-Have Skills 3–5 years of experience in machine learning or applied deep learning roles. Strong Python skills and deep proficiency in PyTorch (or equivalent frameworks). Experience with neural network design, training optimization, and inference tuning. Familiarity with MLOps practices and tools for model deployment and monitoring. Understanding of distributed data processing tools (e.g., Apache Spark, Kafka). Strong grasp of model evaluation, bias/error analysis, and validation workflows. Fluent English. Nice-to-Have Experience with voice/audio data or NLP tasks. Familiarity with vector databases, real-time inference, or edge model deployment. Exposure to infrastructure-as-code tools and scalable cloud environments. Who You Are You turn data into models and models into real-world impact. You care about performance—computational, behavioral, and business—and thrive in fast-moving environments where research meets production. You bring scientific rigor and engineering discipline to every experiment. Why Join Mappa Cutting-edge tech: Tackle voice-AI, real-time inference, and behavioral modeling at once. High ownership: Build and ship models that power our core product. Remote-first: Join a distributed team across LatAm and the U.S.