data scientist/ engineer - Emory, United States - D L RESOURCES PTE LTD

Description

Job Description:

BIG data platform maintenanceCreate, build and maintain the data infrastructure required for Extraction, Transformation and Loading of data from a wide variety of sources.

Ensure high-quality data is delivered to support Compliance / FCC model development. Support the implementation of Data Analytics / Automation for Compliance teams and Business stakeholders.

Liaise with relevant stakeholders (e.g., Data Management Office) to identify and assess suitability of data for Compliance / FCC Data models.

Liaise with Business stakeholders / analysts and ensure assembled data assets used for model development meet business requirements.

Build & maintain data dictionaries / data requirements of implemented data models and coordinate with Data Scientist(s) in developing model narratives.

Monitor the execution & performance of data models and liaise with IT for any technical / system outages. Keep track of datasets / tables in the working schema and support purging of unused tables periodically. Perform data analysis required to troubleshoot data related issues and support the resolution of raised data issues.

Support model maintenanceLiaise with Business Analyst to receive and understand business feedback on model performance and incorporate feedback into models.

Re-train and recalibrate existing FCC analytical models to prevent model drift periodically or as needed.

New model developmentWork closely with model end-users and other key stakeholders (e.g., Head of FCC Analytics, Business Analyst, Data and Ops Engineer, GC) to identify additional areas which require analytics support or future model build and include those models in development pipeline.

Develop model narratives (e.g., purpose, logic, parameters, data requirements, output surfacing / structuring) in collaboration with the business and other relevant stakeholders.

Work closely with other Data Scientist(s) and Business Analyst, and undertake the end-to-end FCC model development, including data wrangling, exploratory data analysis, feature selection, model selection, training, testing, etc.

Build a range of models (rule-based, supervised / unsupervised models, etc.) on structured, semi-structured, and/or unstructured data if needed.#J-18808-Ljbffr