Phani Kumar Email & Phone Number

Harrisburg, Pennsylvania, United States

• Worked with both unstructured/structured data Machine Learning Algorithms such as Linear, Logistic, Decision Tress, Random Forests, Support Vector Machines, Neural Networks, KNN, and Time series analysis.• Automated data collection and cleaning processes, leading to a 50% reduction in data prep time.• worked with Snowflake Schema, data modelling, and its components, as well as source-to-target mappings, interface matrices, and design elements. I have conducted data quality issue analysis using Snow SQL by establishing analytical warehouses on the Snowflake platform.• Conducted hypotheses testing to validate assumptions and guide business strategies, resulting increase in customer engagement.• Used K-means clustering for grouping similar data and documented.• Worked on Numerical optimization, Anomaly Detection and estimation, A/B testing, Statistics, NumPy, SciPy, Pandas, scikit-learn.

California, United States

 • Researched TensorFlow LSTM networks for speech recognition, OpenCV object tracking algorithms, and 11+ new technologies; enhanced 74% of a dancing robot's performance. • Developed interactive dashboards using Data Studio to visualize store performance metrics, such as sales, foot traffic, and customer satisfaction.• A Strong programming experience in developing SQL and PL/SQL codes through various Stored Procedures, Functions, Packages, Cursors, Indexes, Views, and Triggers to implement the business logic of databases.• Researched and evaluated new machine learning technologies for a financial services client, resulting in the adoption of a new algorithm that improved fraud detection rates by 20%.• Built a Chatbot using NLP that increased customer interaction by 70%.• Implemented a Natural Language Processing(NLP) solution that improved RW's chatbot accuracy to 95%, leading to a 63-second reduction in customer service response time. Extracted real-time data streams from various sources, including transactional databases, API endpoints, and event queues, leveraging technologies like Apache Kafka and AWS Kinesis. Leveraged Apache Airflow to orchestrate complex data workflows, scheduling, and monitoring data processing jobs. Integrated Airflow with various data sources and destinations, such as databases, cloud storage, and APIs, to enable seamless data movement and transformation. Collaborated with ETL teams to provide source-to-target data mapping with naming standards, data types, and domain definitions.

India

• Developed and maintained a machine learning pipeline for real-time data processing, resulting in a 50% reduction in data processing time and improved system efficiency.• Improved data mining and automation by 45% using a scalable classifier by integrating predictive models and supervised and unsupervised learning algorithms.• Created 15+ ETL pipeline (to train datasets) and 20+ ML pipelines with PyTorch, ski-kit-learn, and Numpy (training, assessing, and deploying models)• Engineered a Natural Language Processing (NLP) model that increased the speed of text analysis by 60%.• Developed and maintained a software library for a healthcare client, enabling data scientists to build and deploy machine learning models more efficiently and resulting in a 25% reduction in model development time.  Used AWS data pipeline for Data Extraction, Transformation and Loading from homogeneous or heterogeneous data sources and built various graphs for business decision-making using Python matplot library.

India

 • Collaborated with a team of data scientists and engineers to develop a natural language processing (NLP) algorithm for a healthcare client, improving patient diagnosis accuracy by 15% and reducing misdiagnosis rates by 10%.• Worked on Numerical optimization, Anomaly Detection and estimation, A/B testing, Statistics, NumPy, SciPy, Pandas, scikit-learn.• Worked on training the data models, data cleaning and validation in Python.• Developed classification data models in Python by pre-processing the data, cleansing, validating, and testing the data.• Played a pivotal role in data migration projects, enabling smoother transitions to new analytics platforms.• Designed and implemented a machine learning algorithm for a transportation client, reducing delivery times by 20% and improving on-time delivery rates by 15%.• Processed and analyzed sales data, uncovering key trends that led to a 10% lift in sales performance. Developed MapReduce jobs to process and analyze large volumes of structured and unstructured data in parallel, utilizing Hadoop's distributed computing capabilities. Conducted data analysis, data mining, and data engineering to present data clearly and derive actionable insights. Hands-on experience with ETL processes and tools, including data extraction, transformation, and loading, ensuring seamless data integration across systems.