Hi there, I'm
Transforming raw data into actionable insights with 3+ years of experience in machine learning, statistical modeling, and data storytelling.
import pandas as pd
import numpy as np
from sklearn.ensemble import RandomForestClassifier
# Meet Swastik De
profile = {
"role": "Data Scientist",
"experience": "3 years",
"passion": "ML & AI",
"status": "Open to work"
}
print("Building amazing things"
" with data! 🚀")
I'm a passionate Data Scientist with over 3 years of experience turning complex datasets into meaningful insights. I specialize in building end-to-end machine learning pipelines, predictive models, and data-driven solutions that make a real business impact.
My work spans across industries including finance, healthcare, and e-commerce — where I've applied techniques like natural language processing, computer vision, and time-series forecasting to solve real-world problems.
When I'm not wrangling data, I enjoy contributing to open-source projects, writing technical blog posts, and staying up to date with the latest research in AI/ML.
Technologies and tools I work with every day
A selection of my most impactful data science work
Built an end-to-end ML pipeline to predict customer churn for a telecom company. Achieved 92% accuracy using gradient boosting with feature engineering and SHAP explanations. Deployed as a REST API serving real-time predictions.
Developed a multi-class sentiment analysis system for product reviews using BERT fine-tuning. Processed 1M+ reviews with 89% F1-score. Includes an interactive dashboard for real-time sentiment monitoring.
Built a hybrid forecasting model combining ARIMA and LSTM for retail sales prediction. Reduced MAPE by 18% compared to the existing baseline, helping optimize inventory management.
Developed a CNN model using transfer learning (ResNet-50) to classify chest X-rays for pneumonia detection. Achieved 94% AUC-ROC with an explainability layer using Grad-CAM.
Built a collaborative + content-based hybrid recommendation engine for an e-commerce platform. Improved click-through rate by 23% and increased average order value using matrix factorization and embeddings.
Designed a real-time fraud detection pipeline for financial transactions using ensemble models and anomaly detection. Reduced false positive rates by 35% while maintaining high recall.
Tech Company · Full-time
Analytics Firm · Full-time
Startup · Full-time
Indian Statistical Institute · 2021 – 22
Visva-Bharati Central University · 2019-21
Indian Statistical Institute · 2021
I'm currently open to new opportunities. Whether you have a project in mind, want to collaborate, or just say hi — my inbox is always open!