Job Description
We are seeking a talented MLOps Engineer to help operationalize machine learning models that enhance our services and contribute to our mission.
Key Responsibilities:
* Design, build, and maintain scalable ML model deployment pipelines for production environments.
* Develop and manage CI/CD processes tailored for machine learning workflows, ensuring reliable model integration and deployment.
* Implement model monitoring and alerting systems to track model performance metrics (e.g., accuracy, latency) and detect data drift or model decay.
* Collaborate with data scientists to streamline model retraining, versioning, and model lifecycle management.
* Configure and optimize infrastructure for efficient compute resource usage (e.g., GPUs, TPUs), enabling high-performance model training and inference.
* Establish best practices for data and model versioning, experiment tracking, and reproducibility.
* Automate and manage ETL workflows, enabling real-time data availability for training and inference.
* Ensure compliance with data protection regulations (e.g., GDPR) and enforce secure data handling practices.
* Conduct A/B testing and canary releases to assess model performance in production environments.
* Collaborate cross-functionally with software engineers, IT teams, and data scientists to support seamless integration of ML models.
Qualifications:
* Deep knowledge of traditional ML concepts (e.g., LSTMs, RNNs, GMMs, SVMs, trees, boosting) as well as more recent deep learning fundamentals and NLP-related experience with word embeddings.
* Proficiency in JVM languages.
* Familiarity with CI/CD tools and methodologies.
* Proficiency with containerization (e.g., Docker) and orchestration tools.
* Experience with cloud-based ML platforms (e.g., Amazon Sagemaker).
* Experience with common JVM search, linguistics, and other language frameworks (e.g., Lucene, StanfordNLP, OpenNLP, SparkNLP, ANTLR).
* Experience using a Deep Learning Framework (e.g., Tensorflow, PyTorch, Keras).
* Mature theoretical grasp of different neural networks on large-scale datasets.
* Deep and fundamental understanding in signal processing concepts.
* A positive, collaborative, can-do attitude and a strong sense of ownership.
* Familiarity with clinical data, concepts, and language.
* Experience in model training automation with a combination of Supervised, Unsupervised, and Reinforcement methods.