Jyoti Bhandari
West Haven, CT LinkedIn 571-***-**** ad2pn0@r.postjobfree.com Portfolio
EDUCATION
University of New Haven Master of Science in Data Science (GPA: 4.0) Graduation Date: December 2023 West Haven, CT Tribhuvan University B.E. in Computer Engineering (GPA: 4.0) Graduation Date: October 2018 Kathmandu, Nepal SKILLS
• Languages: Python (Numpy, Pandas, PyMc3, Matplotlib, Seaborn, Dash by Plotly, Django, Flask, Scikit-Learn, TensorFlow, Keras, PyTorch) SAS, R, PL/SQL, MongoDB, Linux Bash. Database: PostgreSQL, SQL
• Big data technologies: Spark, Pyspark, Hadoop, Databricks, Apache Airflow, Dagstar, Snowflake
• Tools: SQL Tool, SQL Developer, Bitbucket, JIRA, Jenkins, MicroStrategy, Power BI, Tableau, DataRobot AI Platform, Excel
• Cloud Computing: Amazon Web Services (AWS), EMR, Redshift, Glue, EC2, AWS Lambda, Step Functions
• Soft skills: Effective Communication, Team Collaboration, Adaptability, Problem-Solving, Leadership and Initiative, Time Management
WORK EXPERIENCE
University of New Haven Chief Data Scientist May 2023 – Dec 2023 West Haven, CT
• Reduced admissions application processing time by 25% by developing a model predicting the CGPA, built ML-OPS data pipeline using Apache Airflow with PostgreSQL database, within AWS EC2 instance, and trained Model using Random Forest Regressor generating 82.7% accuracy rate.
• Translated complex machine learning results into accessible presentations, enabling strategic decision-making by leadership.
• Minimized transcript review time by implementing a deep learning model for Optical Character Recognition (OCR), leveraging Document AI which efficiently transforms PDF transcripts into Excel files, preparing them for seamless integration into the CGPA prediction model.
• Improved admission processes by conducting targeted A/B tests to optimize specific aspects of the workflow. These tests, centered on user behaviors and preferences, led to a 15% increase in enrollment rates.
• Managed a cross-functional team of 6 data engineers, 3 data analysts, and 3 data scientists: created work plans and timelines and ensured documentation of codes and changes using version control methods in GitHub to ensure traceability and collaboration resulting in 100% on-time project deliveries.
University of New Haven Data Analyst January 2023 – May 2023 West Haven, CT
• Accomplished a 15% improvement in admission processes' efficiency by developing a causal inference model to understand whether specific changes in the admission process directly contribute to improvements or declines in outcomes.
• Facilitated admissions strategic planning through exploratory data analysis (EDA), feature extraction through correlation analysis, ad-hoc analysis, and hypothesis testing, revealing a notable annual 10% increase in fall semester applications.
• Improved decision-making processes, by creating Power BI dashboards consolidating graduate admissions metrics and KPIs, such as application trends, acceptance rates, and demographic insights, leading to a 20% improvement in application review efficiency. Cotiviti Nepal Pvt. Ltd Software Engineer I (Data Analyst) February 2019 – November 2021 Kathmandu, Nepal
• Improved risk management through data analysis within the insurance and U.S. healthcare domains, focusing on eligibility, claims, prescription claims, and wellness data, measured by a 20% reduction in errors and discrepancies in insurance claims processing, achieved by implementing automated data validation checks.
• Achieved a 20% boost in data ingestion efficiency by leading the design and implementation of streamlined ETL pipelines, using PL/SQL and Oracle Data Integrator (ODI).
• Reduced manual effort by 30% by automating the solutions for identifying gaps, variances, trends, and patterns in data, and expediting insights generation using advanced PL/SQL queries (window functions, procedures, nested queries).
• Reduced the data analysis time by 20% by developing a new report for the team using MicroStrategy and Tableau and proficiently managed the projects using JIRA.
PROJECTS
● Data Science Job Hunt: Built an interactive dashboard using Dash by Plotly to analyze Data Science job opportunities in the US, and predict salaries using Linear Regression, and employee retention using XG Boost Classifier.
● Chatbot Recommending Products: Developed a Discord chatbot using CNN, K-means clustering, KNN algorithm, and VGG16 model to recommend products from the Flipkart database based on visual similarity.
● NutriBot: This Natural Language Processing (NLP) project is a Discord bot, finetuned on Large Language Model BERT, for personalized recipe recommendations and nutritional analysis as per user’s input.
● Walmart Sales Data Analysis: Leveraging advanced time series forecasting methods like ARIMA, a robust predictive model was constructed to enhance Walmart's weekly sales forecasting accuracy by 15%. This data-driven approach optimized inventory management and illuminated temporal trends impacting sales patterns.