Alekhya H Email & Phone Number

Minneapolis, Minnesota, United States

 Collaborated with multiple stakeholders to identify opportunities for leveraging data and analytics to optimize refinery processes. Development and deployment of various batch and real-time Machine Learning systems in Production to optimize the Gross Refinery Margin. Experienced in utilizing distributed data/computing tools such as PySpark to handle big data challenges effectively. Developed data pipelines for ETL (Extract, Transform, Load) processes, automating the flow of data between systems and ensuring data integrity and consistency. Developed and maintained R-based data pipelines for collecting, cleaning, and transforming large datasets from multiple sources. Supported the deployment of machine learning models on Google Cloud AI Platform, ensuring scalability and performance efficiency. Proficient in managing the entire data science project life cycle and actively involved in all phases of the project life cycle.

Boston, Massachusetts, United States

 Creating exception-based email alerts for Proactive communication of actionable insights through pattern recognition and diagnostic investigations with recommendations to prevent future operational issues. Analyze and Prepare data, identify the patterns on the dataset by applying historical models, and collaborate with other Data Scientists to understand data. Perform data manipulation, data preparation, normalization, and predictive modelling. Improve efficiency and accuracy by evaluating models in Python. Implemented data pipelines on Google Cloud Platform (GCP) using services such as BigQuery and Dataflow for efficient data processing and analysis. Contributed to the development of recommendation systems using Google Cloud services, improving personalization and user experience Building ML models like Decision Trees, Regression, Time Series, Boosting and Support Vector Machines using the Jupyter notebook in Seeq Data lab.

Tampa, Florida, United States

 Proficient in SQL for querying and manipulating relational databases and retrieving relevant data for analysis and modelling purposes. Daily monitoring and troubleshooting of the plant and preparing daily and monthly reports for the Operations Team. Skilled in translating ad-hoc reports into comprehensive data analyses, providing actionable insights for strategic decision-making. Proficient in Excel, utilizing advanced formulas and functions to manipulate and analyze large datasets. Expertise in crafting visually appealing presentations using PowerPoint to convey insights effectively. Proficient in Excel, utilizing advanced formulas and functions to manipulate and analyze large datasets.

Hyderabad, Telangana, India

 Provided technical support to clients and internal teams in a Value-Based Care environment, developing data-driven approaches for operational improvement of data infrastructure Migrated data to Snowflake using Fivetran, dbt, and Airflow, managing the data modeling and ETL process Created and optimized scalable data pipeline using python to extract, transform and load structured and semi structured data and designing data models on relational and non-relation databases that can support various business needs including Machine Learning models Successfully performed image processing, OCR, and object detection using Python Tensorflow, Tesseract, OpenCV with over 90% accuracy Performed edge corner detection, image filtering, affine transformation, processed data from an image Forecasted multiple time-series models on big data with Python, Spark, and Databricks using fb Prophet, achieving a 99% reduction in execution time from 61.7 to 0.82 seconds Designed and implemented data pipelines and ETL (Extract, Transform, Load) processes using Spark Optimised Spark jobs for performance, such as through partitioning, caching, and tuning various configuration parameters, to ensure efficient data processing and reduce execution times Collaborated with Data Scientist to develop and evaluate multiple ML models (SVM, Naïve Bayes, Random Forest) for multi-class text classification, compared models using cross-validation to identify the optimal one, classified with 98% accuracy using linear SVM.