About Me.
I'm Jenil Shah, a curious mind with a passion for solving puzzles – whether on a chessboard, in the kitchen, or in the complex world of data. Born in India and now calling Boston home, I've transformed challenges into opportunities – from being a state-level swimmer to becoming a data analytics professional.
At 17, I took a leap of faith and moved to the U.S., transforming challenges into opportunities. When I'm not diving deep into data, you'll find me strategizing on a chess board, experimenting with new recipes in the kitchen, or planning my next travel adventure. These experiences have taught me that problem-solving is an art, whether it's in data analysis or life.
Life, to me, is about continuous learning, embracing the unexpected, and finding joy in the journey of discovery.
Tools I Know.
-
Python
-
R
-
MySQL
-
BigQuery
-
Snowflake
-
AWS
-
GCP
-
Tableau
-
MS Excel
-
Git
-
Docker
-
Kubernetes
-
Apache Spark
-
Hadoop
-
Streamlit
-
SAS
-
Databricks
My Work Experience.
Throughout my career, I've had the privilege of working with diverse organizations, each contributing significantly to my professional growth and expertise as a data professional. Each of these experiences has been instrumental in shaping my expertise, from technical proficiency to strategic thinking. They've equipped me with a versatile skill set and a holistic understanding of how data can drive business success across different industries and contexts. Here's a snapshot of my journey:
Board
Professional Services (Analytics Consultant) Intern
May 2024 - Dec 2024- Led data migration efforts from Board version 12.6 to 14.1, conducting thorough regression testing on multiple client applications, resulting in a 25% reduction in post-upgrade issues and improvements in application performance, ensuring seamless functionality and consistent UI/UX for clients.
- Developed a comprehensive financial planning and analysis demo for Nvidia and presented it to the Board team, integrating budgeting and reporting functionalities, resulting in a 40% reduction in data processing time.
- Created essential Standard Operating Procedure (SOP) documents for new user creation, cloud admin portal, and transporter tool, resulting in a 30% reduction in client upgrade process time and a 40% decrease in user-related support tickets for the North American team during migrations.
Boston University
Teaching Assistant: Operations and Supply Chain Management
Jan 2024 - May 2024- Supported 50+ students in mastering supply chain analytics tools and techniques in OM323 class. Facilitated weekly lab sessions and held office hours to provide hands-on guidance with software applications and complex supply chain concepts.
- Graded assignments and exams, offering constructive feedback to enhance students' analytical and problem-solving skills. Developed detailed rubrics and provided personalized comments to students.
- Collaborated with professor to develop case studies based on current supply chain disruptions and resilience strategies. Researched real-world supply chain challenges and crafted engaging scenarios that allowed students to apply theoretical knowledge to practical situations.
Ergode Inc.
Business Data Analyst
Sep 2021 - Aug 2022- Developed python scripts to automate data extraction and dashboards to model inventory levels leading to a reduction in processing time by 50% and excess inventory by 20% respectively.
- Built and maintained KPI dashboards with Tableau, creating interactive visualizations that provided real-time business insights, reducing decision-making time by 50%.
- Spearheaded cross-functional analysis for eco-friendly product launch strategy, resulting in successful implementation and 15% revenue increase within the first quarter.
Hitesh Steel Syndicate India Private Limited
Data Analyst
Apr 2020 - Mar 2021- Conducted statistical analysis on TMT rebars sales data and identified significant seasonal patterns, resulting in an optimized inventory management strategy for the client.
- Analyzed industry trends and identified gaps in the market, leading to the development of a new product line that generated $500K in revenue within six months.
- Collaborated with the sales team to create customized dashboards that provided insights into individual sales performance, leading to a 25% increase in team productivity.
My Projects.
My portfolio showcases a diverse range of projects where data analytics drives tangible results. From predictive modeling to real-time dashboards, each endeavor demonstrates my ability to transform complex data into actionable insights. These projects span various industries, tackling challenges in customer retention, recommendation systems, fraud detections, operational efficiency, and strategic decision-making.
By leveraging advanced analytics techniques and cutting-edge technologies, I've consistently delivered solutions that not only solve immediate problems but also create long-term value. This collection of work reflects my passion for using data to uncover opportunities, optimize processes, and fuel innovation in today's data-driven world.
Netflix Recommendation Analytical Pipeline
- Designed a robust ETL pipeline using Prefect and Google Cloud Storage, integrating MotherDuck for efficient data querying and enabling seamless model metric tracking.
- Implemented scalable KNN recommendation models deployed via Google Cloud Functions and Cloud Run, leveraging cosine similarity for precise, real-time Netflix content suggestions
- Built an innovative user-centric Streamlit application enhanced with Gemini 1.5 Pro LLM, enabling personalized AI-driven recommendations secured through Google Secret Cloud Manager.
Streamlit
Job Listing Fraud Detection System
- Implemented a machine learning model to detect fraudulent job listings, achieving 87.4% balanced accuracy using Natural Language Processing techniques like TF-IDF, Word2Vec, and PCA for dimensionality reduction.
- Addressed class imbalance with SMOTE and implemented advanced classification algorithms (Logistic Regression, Random Forest, SVC, XGBoost) and a Stacking Classifier to enhance model performance.
- Engineered diverse feature set from job listings, incorporating text analysis, salary normalization, and geographical clustering, boosting fraud detection model precision and recall by 15%.
Steam Player Retention and Review Analysis
- Developed complex SQL queries in BigQuery to clean, filter, and aggregate data from a multi-tabled steam dataset, reducing processing time by 40% and allowing for more timely analysis.
- Designed interactive dashboards in Tableau to visualize user behavior patterns, pricing dynamics, and game popularity factors, resulting in a 30% increase in stakeholder understanding of key data insights.
- Adapted quickly to changing project requirements, utilizing strong problem-solving skills to identify alternative solutions and improve efficiency, resulting in a 15% decrease in project timeline.
Tableau
London Cycle Hire Investigation (Big Data)
- Used Hadoop and PySpark to analyze usage patterns across demographics and zones, improving resource allocation by 20%.
- Utilized Hive and SQL to extract insights from historical data, increasing service quality by 15%.
- Leveraged Python and PySpark to process 10TB of urban mobility data, developing predictive models that improved infrastructure planning efficiency by 25% and reduced traffic congestion in key areas by 15%.
London Airbnb Price Prediction Analysis
- Employed Python libraries including pandas and seaborn for data pre-processing, exploration and visualization, identifying critical trends and factors within the dataset using modelling decisions.
- Utilized supervised machine learning methods including Random Forest, Linear Regression, KNN, and Support Vector Regressors to construct predictive models, selecting the optimal model for Airbnb price prediction.
- Conducted an in-depth analysis using hyperparameter tuning techniques to optimize model performance, resulting in a 15% increase in accuracy for the predictive model.
Asset Intensity Prediction with Text Analysis using BERT
- Explored the feasibility of predicting a company's asset intensity (debt-to-equity, price-to-sales, return-on-assets) using textual descriptions and BERT for context-aware analysis.
- Built logistic regression models achieving up to 88.7% accuracy in predicting debt-to-equity ratio, demonstrating potential for text analysis in financial assessment.
- Highlighted the need for further exploration for price-to-sales and return-on-assets ratios, suggesting text descriptions may not be the sole indicator for these metrics.