Dhwani Pranav Tank

Results-driven Data Engineer and Machine Learning Associate with 2.5 years of experience in designing scalable data pipelines and developing machine learning models. Specialized in ETL pipeline design, multimodal data processing, and training models for production. Skilled in Python, SQL, AWS, Databricks, and data visualization tools such as Tableau and Power BI.

• VAST Challenge MC2, Visual Analytics, Tableau, D3.js, Identified anomalies in 138,824 rows of waterway data using OpenRefine, proposed improvements in sampling methods, and visualized findings with D3.js and Tableau., Created a radial chart with D3.js to track 112 chemicals and highlight 12 toxic variants, facilitating better understanding and decision-making.
• Big Data Project Management and Visualization, Python, Redis, Tableau, Data Visualization, Video, Managed a 100,000-row dataset with Redis, implementing caching strategies and optimizing data structures for enhanced performance and real-time access., Created interactive dashboards and visualizations in Tableau, transforming complex datasets into actionable insights to support strategic decision-making.
• Data Warehouse Integration and Business Intelligence, Python, ETL, Data Profiling, Directed the calculation, analysis, and consolidation of yearly sales data, ensuring accurate and comprehensive reporting while optimizing data integrity through profiling., Oversaw the ETL process for integrating data from multiple sources into a data warehouse, designed fact tables, and collaborated with teams to align data for reporting.
• Health Data Analysis and Predictive Model, Python, EDA, Statistical Methods, Visualization, Selected and cleaned a health dataset, conducted exploratory data analysis (EDA) in Python, and applied statistical methods to uncover trends related to heart disease., Utilized machine learning techniques to build accurate heart disease models, leveraging insights from data analysis and visualizations to enhance performance.
• Document-Based Question Answering System with LLM-Driven Retrieval-Augmented Generation, Translation, and Sentiment Analysis, Project, Developed a document-based QA system that processes various file formats (PDF, DOCX, TXT), extracts relevant content using LangChain, and generates contextual answers through the Retrieval-Augmented Generation (RAG) model with Sentence Transformer and FAISS., Integrated multi-language support for French, Spanish, and German translations, along with sentiment analysis to assess response tone, providing users with a scalable, user-friendly solution for extracting insights from complex documents.