Siva

About Siva

I have extensive experience in data science, machine learning, and backend development…

Experience

1. Data Scientist

   Johnson & Johnson

   Apr 2023 — Present · Raritan, NJ, US

   I specialize in processing diverse data formats (JSON, XML, CSV) and applying machine learning algorithms using TensorFlow, Scikit-Learn, and PyTorch for predictive modeling and insights extraction. I automated CSV-to-JSON conversion for chatbot integration through NLP scripts, reducing development time by 20% and streamlining data pipelines. Additionally, I migrated SQL Server databases to Azure Cloud, enabling scalable data warehousing and predictive analytics with Azure Machine Learning, improving data accessibility and reducing query response times by 30%.Leveraging Azure Databricks for large-scale data processing, I built ETL pipelines using PySpark and Spark SQL, reducing processing time by 40%. My experience includes utilizing Spark, Hadoop, and Kafka for real-time data processing, increasing user lifetime value by 45%. I engineered a hybrid NLP model combining LSTM and Word2Vec, improving incident ticket categorization accuracy by 35%. I also fine-tuned LLMs like BERT and GPT for conversational AI and sentiment analysis, achieving a 15% accuracy improvement in domain-specific tasks.I designed and deployed Retrieval-Augmented Generation (RAG) pipelines for domain-specific knowledge retrieval, improving response accuracy by 25%. By integrating RAG with LLMs, I enabled context-aware question answering and knowledge extraction. I developed AI agents using transformers and LLMs to automate workflows, monitor forecasting insights, and drive real-time decision-making, increasing user retention by 45% and tripling engagement in target categories.