Professional Contributions
Highlights
- Performed feature selection on 150+ features using XGBoost and Cancelout, increasing the true positives by ~4%.
- Automated capturing logs from Huawei MLOps platform.
- Implemented early stop on Deep and Cross Network-based models, increasing the true positives by ~6%.
- Automated data preprocessing using Dataiku for Citibank, reducing the load by ~45%.
Responsibilities
- Identified the importance of 150+ features using manual and ensemble-based feature selection methods for Huawei Client.
- Implemented early stop and performed hyper-parameter tuning manually for advertisement types, observing significant improvement on real-time data for Huawei Client.
- Played a major role in automating KPI capture from logs, addition/removal of features using Python scripts for Huawei Client.
- Performed data cleaning and preprocessing of customer and beneficiary databases using the Dataiku platform for Citi Bank Client.
Client Journey
-
Huawei Technologies
Optimizing a deep leanring model for Cost Per Click/Impression Prediction
-
Oil and Gas Company
Automated preprocessing workflows in Dataiku, reducing data preparation time and improving data quality.
Certifications and Accreditations
- Recognized as ‘Outstanding Contributor’ in February 2022 by the client (Huawei).
- Completed RapidMiner Machine Learning Professional Certification.
- Completed RapidMiner Machine Learning Master Certification.
- Completed ML Practitioner Certification by Dataiku.
- Completed Core Designer Certification by Dataiku.
- Completed Advanced Designer Certification by Dataiku.