Migrated business logic from PL/SQL to PySpark improving scalability and performance of data processes. Automated deployment of PySpark notebooks using Jenkins CI/CD. Validated and reconciled historical data with data steward ensuring data integrity and consistency. Optimized PySpark processes and refactored code to improve efficiency and reduce technical debt by correcting bad practices.

Developed Scala Jars for data ingestion, curation, and exploitation. Created process to extract CSV files from SFTP server to ADLS and load into Delta tables orchestrated in Databricks workflow. Configured Aercorsoft tasks for SAP table ingestion to ADLS in parquet format. Developed historical load process to Delta tables using liquid clustering. Merged CDC records with historical tables. Transformed data per business rules using PySpark. Orchestrated stored procedures and report exports with Airflow. Optimized BigQuery queries to reduce data retrieval time in PowerBI. Developed clustering models for customer churn using K Means and DBSCAN algorithms.

Ingested data from Oracle DB to BigQuery using Apache Beam and DataFlow. Migrated Oracle PL/SQL procedures to BigQuery. Standardized and homogenized data. Developed orchestration mesh for processes via Cloud Composer and Airflow. Developed ETL using DataFlow Apache Beam.

Implemented a vehicle license plate recognition system using computer vision and deep learning techniques.

Extracted, cleaned, analyzed, and visualized features from seismographs. Implemented algorithms like SMOTE and cost-sensitive methods to alleviate class imbalance. Classified seismic signals from Sabancaya volcano using ensemble of deep neural networks and random forests. Developed a FastAPI for batch predictions.

Designed a regularization algorithm to mitigate overfitting in deep neural networks under data scarcity for regression and classification tasks.

Developed a mobile application presenting real-time data from a galvanic sensor with interactive dashboards to analyze human skin response to stimuli.