About the job
Role Name: Data Scientist
Experience Range: 5 to 14 Years
Location of Requirement: Mumbai/Pune/Kolkata
Job Description
Lead and own the end-to-end lifecycle mgmt. of ML models ensuring high availability, performance & quality of deployed models –
(i) Work with AI/ML architect to build standard ML Pipelines to enable rapid operationalization of ML models with consistent and standard ML performance health and quality monitoring techniques auto enabled.
(ii) Automate the deployment and management of machine learning models using CI/CD practices and tools.
(iii) Monitoring the performance of ML models deployed in production (using tools such as ELK, cloud watch, Graphana) and build automated ML monitoring capabilities for model performance health, quality, drift monitoring in a consistent and automated fashion and trigger model retraining jobs/requests when the performance drift is significant.
(iv) Manage infrastructure for machine learning models in a reproducible and scalable manner, using infrastructure as code practices.
(v) Review the performance of the model & Identifying ways to improve model performance, availability and robustness.
(vi) Maintaining machine learning models in production, including updating them with new data and retraining as needed, and troubleshooting any issues that arise, perform RCA and coordinate with ML engineers and data scientist for model rebuild/retraining.
(vii) work effectively with other team members, including data scientists, cloud devOps engineers, etc. to ensure the smooth and efficient operation of machine learning models in production
Skill Set:
Python Knowledge
Knowledge of CI/CD tools to automate ML pipeline mgmt. (Jenkins/AWS code pipeline)
Experience in AWS cloud native services (including Sage maker, infra., data transformation, cloud monitoring)
Working understanding of a variety of machine learning algorithms, including supervised learning algorithms, unsupervised learning algorithms, and deep learning algorithms
You can share your updated Cv @ [email protected]