I am Sanman Kadam, an MSc Statistics student and aspiring Data Analyst based in Mumbai, India.
My work focuses on transforming raw datasets into clear, decision-ready insights using Python, SQL, Power BI, Excel, and statistical analysis. I have worked with government and business datasets involving data cleaning, exploratory data analysis, KPI reporting, machine learning, dashboarding, NLP classification, churn analysis, forecasting, and policy-level reporting.
I build analytics workflows that move from messy data to structured output: cleaning, validation, feature engineering, model development, evaluation, visualization, and business interpretation. My strongest areas are applied statistics, data analytics, BI dashboards, machine learning, and practical reporting.
| Open To | Focus Areas |
|---|---|
| Data Analyst Roles | SQL, Python, Power BI, Excel |
| BI Analyst Roles | Dashboards, KPI Reporting, DAX |
| Junior ML / Data Science Roles | Classification, Regression, Forecasting |
| Research Analytics Roles | Statistical Testing, EDA, Policy Data |
| Domain | Proficiency | Details |
|---|---|---|
| Exploratory Data Analysis | Advanced | Data cleaning, missing-value review, descriptive statistics, outlier checks, business interpretation |
| Machine Learning | Intermediate | Regression, classification, model evaluation, feature engineering, cross-validation |
| NLP | Intermediate | Text preprocessing, TF-IDF, sentiment analysis, spam classification, resume screening workflows |
| Deep Learning | �Intermediate | LSTM and ANN-based forecasting/classification projects |
| Business Intelligence | Advanced | Power BI dashboards, Power Query, DAX, KPI analysis, stakeholder reporting |
| Statistical Analysis | Advanced | Hypothesis testing, regression, A/B testing, model comparison, inference |
| Forecasting | Intermediate | Lag-based features, time-series modeling, error metric comparison |
Cat Risk A/B Testing
Insurance analytics project comparing flat pricing and risk-based pricing under catastrophe risk using A/B testing, Monte Carlo simulation, VaR, TVaR, and Streamlit.
| Field | Details |
|---|---|
| Stack | Python, Jupyter Notebook, Streamlit |
| Scale | Insurance pricing simulation and risk comparison |
| Performance | Risk-based pricing evaluated using catastrophe-risk metrics |
| Security | No sensitive production data included |
| Impact | Demonstrates actuarial analytics, experimentation, and risk modeling |
| Repository | cat-risk-ab-testing |
This project is a strong analytics portfolio piece because it combines business experimentation with risk modeling. It shows the ability to compare pricing strategies, simulate uncertainty, and explain risk-adjusted decisions using statistical thinking.
E-Commerce Conversion Optimization
A/B testing project for checkout conversion optimization using z-test, logistic regression, Bayesian analysis, and business impact estimation.
| Field | Details |
|---|---|
| Stack | Python, Statistics, Logistic Regression |
| Scale | Checkout conversion experiment |
| Performance | Evaluates uplift and launch/no-launch decision logic |
| Security | Uses project-level analysis data only |
| Impact | Converts experiment results into business recommendations |
| Repository | conversion-optimization-ab-testing |
This project shows practical experimentation skill: defining treatment impact, handling confounders, comparing statistical approaches, and translating model output into a decision.
Retail Sales Analytics Dashboard
End-to-end retail sales analysis project using Python, SQL, and Power BI.
| Field | Details |
|---|---|
| Stack | Python, SQL, Power BI |
| Scale | Retail sales KPI reporting |
| Performance | Revenue, AOV, customer segmentation, and product-level insights |
| Security | No confidential business data exposed |
| Impact | Turns raw sales data into dashboard-ready business intelligence |
| Repository | retail-sales-analytics-dashboard |
This project demonstrates the full analytics workflow: cleaning, querying, KPI design, dashboarding, and interpretation. It is built for business users who need clear insights rather than raw tables.
COVID-19 Time Series Forecasting
Time-series forecasting of COVID-19 daily new cases using Linear Regression, ANN, and LSTM with lag-based features.
| Field | Details |
|---|---|
| Stack | Python, Machine Learning, ANN, LSTM |
| Scale | Daily case forecasting |
| Performance | Compared models using RMSE, MAE, and R² |
| Security | Public health dataset style workflow |
| Impact | Shows applied forecasting and model comparison |
| Repository | covid19-time-series-forecasting |
This project highlights the ability to build lag-based forecasting datasets, compare classical and neural models, and select the best approach using error metrics.
SMS Spam Classification ML / DL
SMS spam detection using machine learning models and Bidirectional LSTM with strong classification performance.
| Field | Details |
|---|---|
| Stack | Python, NLP, Machine Learning, LSTM |
| Scale | Text classification pipeline |
| Performance | Reported accuracy: 98.16% |
| Security | Uses non-production text classification data |
| Impact | Demonstrates NLP preprocessing, classification, and model evaluation |
| Repository | SMS_Spam_Classification_ML_DL |
This project is useful for demonstrating NLP and classification fundamentals: text preprocessing, feature extraction, model comparison, deep learning, and performance reporting.
Regularized Regression Study
Comparative analysis of OLS, Ridge, and Lasso regression with cross-validation, regularization tuning, and feature selection.
| Field | Details |
|---|---|
| Stack | Python, Scikit-learn, Statistics |
| Scale | Regression modeling study |
| Performance | Compares regularized models against baseline regression |
| Security | No sensitive data included |
| Impact | Shows statistical modeling discipline and overfitting control |
| Repository | regularized-regression-mtcars |
This project reflects strong statistical modeling habits: comparing baseline and regularized models, tuning hyperparameters, and interpreting feature selection.
Credit Default Prediction
Machine learning project using SVM to predict credit card default risk with preprocessing, class balancing, hyperparameter tuning, and PCA visualization.
| Field | Details |
|---|---|
| Stack | Python, SVM, PCA, Machine Learning |
| Scale | Credit risk classification |
| Performance | Model tuning and class-balance workflow |
| Security | No live financial data exposed |
| Impact | Shows applied risk classification and model evaluation |
| Repository | svm-credit-default-prediction |
This project demonstrates classification modeling for risk analytics, including preprocessing, balancing, tuning, and dimensionality reduction.
AI Emotional Research Chatbot
Chatbot-based data collection system for research on emotional interaction with AI.
| Field | Details |
|---|---|
| Stack | Python |
| Scale | Research data collection workflow |
| Performance | Supports structured survey-style interaction |
| Security | Research-purpose project repository |
| Impact | Connects data collection, user interaction, and research analytics |
| Repository | ai_emotional_research_chatbot |
This project shows practical implementation of a research-support tool, connecting Python development with survey data collection and applied analytics.
EVOASTRA VENTURES PVT LTD, Mumbai
Oct 2025-Nov 2025
Worked on an end-to-end telecom churn prediction workflow using Python and Scikit-learn. The role involved exploratory data analysis, feature engineering, web scraping automation, and model evaluation.
- Developed churn prediction workflow using Python and Scikit-learn.
- Performed EDA to identify churn drivers and behavioral patterns.
- Engineered features for model-ready datasets.
- Built automated web scraping pipeline using BeautifulSoup and Selenium.
- Evaluated models using accuracy and F1-score.
Directorate of Economics and Statistics, Mumbai
May 2024-Jun 2024
Worked on government datasets involving cleaning, consolidation, dashboarding, and policy-level reporting. The role focused on making public datasets more consistent and usable for analysis.
- Cleaned and consolidated 100K+ government records using Excel and Python.
- Built interactive Power BI dashboards using Power Query and DAX.
- Analyzed public datasets on water, sanitation, and housing.
- Supported regional planning decisions through structured reporting.
- Improved consistency in policy-level data outputs.
| Recognition | Details |
|---|---|
| Best Group Leader Trophy | Awarded for NSS leadership and community initiatives |
| Government Data Analytics Experience | Worked with 100K+ public-sector records during DES internship |
| Applied Analytics Portfolio | Built public projects across BI, ML, NLP, forecasting, A/B testing, and risk analytics |
| MSc Statistics Background | Strong academic foundation in statistics, inference, regression, and analytics |
Learning:
- Advanced SQL for analytics
- Power BI dashboard design
- Machine learning model evaluation
- Statistical modeling for business decisions
Building:
- Data analytics portfolio projects
- BI dashboards with KPI storytelling
- NLP and classification workflows
- Forecasting and experimentation projects
Exploring:
- A/B testing and causal inference
- Risk analytics
- Research data collection systems
- Applied statistics in real-world datasets
Open To:
- Data Analyst roles
- BI Analyst roles
- Junior Data Scientist roles
- Research analytics opportunities
