At Roche, the business problem is to prevent, stop and cure diseases and ensure everyone has access to healthcare today and for generations to come, by leveraging machine learning, data science, and software engineering.
Requirements
- Proficiency in machine learning algorithms, including traditional models and deep learning architectures
- Experience with Natural Language Processing (NLP) techniques, including tokenization, sentiment analysis, and named entity recognition
- Strong data handling and feature engineering skills, including data cleaning, preprocessing, and transformation
- Proficiency in Python, along with key libraries like Scikit-learn, Pandas, TensorFlow, and PyTorch
- Experience with MLOps practices, including model versioning, monitoring, and deployment on cloud platforms
- Understanding of AI agent architectures, including Large Language Models (LLMs) and logical structures for decision-making
- Experience with workflow automation and handling unstructured data, including text, images, and audio
Responsibilities
- Machine Learning and Deep Learning: The candidate must be proficient in a wide range of ML algorithms, from traditional models like linear regression and decision trees to more advanced deep learning architectures such as Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs).
- Natural Language Processing (NLP): For extracting information from unstructured text, strong NLP skills are essential. Look for experience with techniques like tokenization, sentiment analysis, named entity recognition, topic modeling, and using pre-trained language models like BERT, GPT, or others from the Hugging Face ecosystem.
- Data Handling and Feature Engineering: They should be adept at working with various data formats and have experience in data cleaning, preprocessing, and transforming raw data into useful features for ML models.
- Programming and MLOps: Proficiency in Python is a must, along with a solid understanding of key libraries like Scikit-learn, Pandas, TensorFlow, and PyTorch.
- AI Agent Architectures: Look for a candidate who understands the components of an AI agent, including a Large Language Model (LLM) as the brain, tools for specific tasks, and a logical structure for decision-making.
- Workflow Automation: The candidate should have practical experience in designing and implementing automated workflows.
- Unstructured Data: The candidate needs to demonstrate expertise in handling various forms of unstructured data, including text, images, and audio.
Other
- Problem-Solving: The ability to break down complex business problems into manageable, data-driven solutions
- Communication: A great candidate can clearly articulate technical concepts to non-technical stakeholders
- Business Acumen: The best candidates understand the business context of their work and can connect their technical solutions to a positive impact on the company's bottom line or operational efficiency
- Relocation benefits are not available for this posting
- A discretionary annual bonus may be available based on individual and Company performance
