SHARAD VASANT PARMAR

**** ********* ****, *** ***, San Diego CA-92115 ad3405@r.postjobfree.com 619-***-**** LinkedIn EDUCATION

San Diego State University, San Diego, California Aug 2023 – May 2025 Masters in Big Data Analytics CGPA: 3.6/4.0

Relevant Coursework: Principles of Data Science, Implementation of Machine Learning, Advanced Mathematical Statistics Dwarkadas J Sanghvi College of Engineering, Mumbai, India May 2018 – Jun 2022 Bachelor of Technology in Electronics and Telecommunication CGPA: 3.8/4.0 Data Science Graduation Credited certification course (30 credits) with IBM ICE CGPA: 4.0/4.0 SKILLS

• Programming Language: Python (scikit-learn, pandas, numPy, NLTK), R

• Data Visualization: Python(matplotlib/seaborn/plotly/dash), Tableau

• Database: SQL (joins, group by, sub queries, window functions, CTE), MongoDB, Neo4j

• Statistical tool: Stata, JMP, Excel (pivot table, Optimization)

• Specialization: Statistical Analysis, Time Series Analysis, Fraud Analysis, Text Analytics, Machine Learning, A/B testing, Hypothesis Testing, Dashboarding

WORK EXPERIENCE

SDSU Strategic Communications Department, San Diego, CA Sep 2023 – Apr 2024 Data Science Researcher

• Utilized machine learning techniques including classification and clustering, to analyze audience engagement patterns, refining strategies accordingly and achieving a notable 25% boost in interaction rates.

• Implemented Python, SPSS, and SAS for thorough data analysis, driving a substantial 3x improvement in outreach effectiveness. Presented findings succinctly using Looker Studio.

• Engineered predictive models to anticipate audience behaviors, facilitating prompt adjustments in communication tactics and yielding a measurable 15% increase in the university's brand visibility. Think Analytics Pvt ltd, Mumbai, India Jul 2022 – Aug 2023 Data Scientist (Client Stolt Neilson - Carbon Emissions Forecasting)

• Developed GLM predictive model for Carbon emissions forecasting using L1 and L2 regularization, launching POC project to production, achieving $3M annual cost reduction in large-scale data-driven decision-making.

• Implemented a cloud-based solution using iterative modeling in Azure Machine Learning for forecasting fuel consumption across 187 vessels, aiding business stakeholders and saving $1M for every "D" and "E" rated vessels.

• Utilized SQL Server Management Studio to transform GBs of live data from daily to Voyage level. Implemented pipeline for linear model using Azure ML, resulting in a process 3x faster and fully automated.

• Generated and presented detailed weekly reports to clients, leveraging tools such as Python, Power BI, and data visualization libraries (including Matplotlib, Seaborn), enabling data-driven business decisions.

• Leveraging technical expertise and strategic alignment, facilitated seamless interaction with Business Analysts and stakeholders, resulting in the harmonization of data science goals with overarching business objectives of the project. Think Analytics Pvt Ltd, Mumbai, India Aug 2021 – Feb 2022 Data Science Intern (Client Stolt Neilson- Carbon Emissions Forecasting)

• Developed ARIMA and gradient boosting models in Python using SciKit-Learn and StatsModels for vessel Carbon Intensity Indicator time series forecasting, aiming for $0.8M in annual savings per vessel.

• Utilized H2O.ai Auto-ML in Python led to the selection of the optimal supervised machine learning model, yielding an 18% validated reduction in fuel consumption, integrated model results into a user-friendly interface using Streamlit. PERSONAL PROJECTS

Healthcare Real Time Data Analysis San Diego State University Oct 2023 - Dec 2023

• Designed and implemented an Extract, Transform, Load (ETL) pipeline optimized for real-time healthcare data. The pipeline ensured smooth integration of healthcare data, enhancing the overall efficiency of the system.

• Leveraged AWS services such as AWS Glue for automated ETL processes. This automated data transformation ensured accuracy and consistency while securely storing data in AWS S3, complying with data security standards. Stock Price Prediction Using Time-Series Forecasting

• Developed a Bi-LSTM-based DL model for time series analysis, predicting stock values with Apple Inc. data.

• Integrated 2015-2018 stock prices, achieving 86.8% accuracy in forecasting 2019 stock prices.

Contact this candidate