MLOPs Engineer

Key Skills

Aartificial intelligence, AWS, Continuous Integration & Continuous Delivery - CI/CD, Data science techniques, Docker, GitLab, Kubernetes-K8s, Linux Operating system, Machine learning techniques, MLOps tools, Optimization, PyTorch, TensorFlow

Job Description

Roche has established Global Analytics and Technology Center of Excellence (GATE) to drive analytics & technology driven solutions by partnering with Roche affiliates across the globe. While the candidate can join us remotely, we would require them to move to our Chennai or Hyderabad Office soon. To know more about us, visit - https://www.rocheindia.com/en/About_Roche/GATE.html

The MLOps Engineer will play a crucial role in ensuring the accuracy, performance, and stability of machine learning models used for forecasting and analytics. This role involves continuous monitoring, maintenance, and improvement of ML pipelines, Docker images, and data synchronization processes to support reliable predictions and decision-making

An ideal candidate should have:

• Overall experience of 4-6 Years, with experience in MLOps space is a must
• Design and implement cloud solutions, build MLOps on cloud (AWS, Azure, or GCP),
• Build CI/CD pipelines orchestration by GitLab CI, GitHub Actions, Circle CI, Airflow or similar tools;
• Data science model review, run the code refactoring and optimization, containerization, deployment, versioning, and monitoring of its quality
• Data science models testing, validation and tests automation
• Programming languages like Python, Go, Ruby or Bash, good understanding of Linux, knowledge of frameworks such as scikit-learn, Keras, PyTorch, Tensorflow, etc.
• General understanding of AGILE Project Delivery process
• Excellent written and verbal communication skills for coordinating across teams
• Strong analytic mindset and logical thinking capability, strong QC mindset
• Demonstrates consulting, creativity, critical thinking, project planning, and attention to detail capabilities
• Knowledge of US/Europe pharmaceutical market and experience with pharmaceutical data would be a plus, but not a must

As an MLOPs Engineer, you will do:

1) Accuracy Analysis:

• Perform in-depth analysis of forecasting model accuracy
• Identify and diagnose issues causing low accuracy
• Collaborate with data scientists to understand and rectify model shortcomings

2) Performance Comparison:

• Regularly compare model performance against the previous month's results
• Identify trends and areas for improvement in forecasting accuracy

3) Pipeline and Docker Maintenance:

• Ensure that the production pipeline and Docker images are running with the latest model code
• Perform regular checks to maintain the stability of the production environment

4) Development-Production Alignment:

• Align the development environment with the production environment every two weeks to ensure consistency

5) Bug Fixing:

• Investigate and fix code bugs related to model deployment and data synchronization

6) Data Synchronization:

• Maintain data synchronization processes to ensure data consistency across systems

7) Accuracy Comparison:

• Implement accuracy comparison between past forecasts and realized actuals to assess model performance

8) Affiliate Support:

• Collaborate with affiliate teams to understand and address their needs related to predictions and model outputs

9) Understanding Predictions:

• Provide insights and explanations for model predictions to relevant stakeholders

10) Model Performance Deep Dive:

• Conduct in-depth analysis of model performance, including feature importance and model behavior

11) Monthly Export:

• Export results in a wide format for monthly reporting and analysis

12) Eventing:

• Upload and process monthly CSV data into the database
• Implement eventing logic to derive insights from data
• Continuously improve eventing code logic for better results

Education:

• UG: B.Tech/B.E. - Any Specialization (Computer Science preferred)
• Master’s Degree – Data Science, Machine Learning, Statistics, Economics/Econometrics, Computer Science or related field- Nice to have
• Any certifications on AI/ML/Data science would be an added advantage