I'm Einstein (Currently open to new roles) - Résumé PDF | Cover Letter
- I specialize in Statistical Inference, with strong expertise in Propensity Modeling, Time Series Forecasting, and practical applications of Generative AI.
- Skilled in using data to drive insights, especially in churn prediction, conversion optimization, demand forecasting, and customer segmentation.
- Always looking to collaborate on data science projects involving machine learning, NLP, and applied statistics.
- Reach me: [email protected]
📌 Propensity Modeling
Using logistic regression, tree-based models (XGBoost, LightGBM), and calibrated classifiers to estimate probabilities of user actions such as:
- Churn risk prediction.
- A/B test targeting for uplift modelling.
- Targeted marketing response likelihood.
- Purchase or conversion (e.g., subscription to financial products).
📌 Statistics for Real-World Impact
- Model calibration (e.g., Platt scaling, isotonic regression).
- Hypothesis testing & confidence intervals for campaign performance.
- Causal inference (e.g., Average Treatment Effect for experiment design).
- Counterfactual analysis to estimate what would have happened under different circumstances.
📌 Time Series Forecasting & Optimisation
- Classical models: ARIMA, SARIMA, and Exponential Smoothing.
- Feature engineering with lag variables, rolling stats, and calendar effects.
- Forecasting for metrics like demand, churn over time, marketing ROI, etc.
- ML/Deep Learning approaches: Facebook Prophet, Neural Prophet, XGBoost, LSTM.
- Parameter tuning via grid search/CV, AIC/BIC selection for AR terms (AR, MA, ARMA).
- Working with additive & multiplicative decompositions to model seasonal, trend, and irregular components.
📌 GenAI & LLMs
- Smart chatbots for lead qualification & feedback collection
- Summarisation of long-form transcripts (e.g., meetings, user calls)
- RAG systems using LangChain + OpenAI for context-aware insights
- Publish a research paper in applied ML or statistics
- Help others be more comfortable with statistics by building projects on uplift modelling, churn prediction and more while talking about them online and in person
- I love making complex topics in ML/stats feel intuitive and practical
- I usually code with music on, unless debugging some stubborn issue 😅
- I believe in learning out loud and sharing simplified insights with the community
- Embeddings in AI & Machine Learning: A Deep Dive into Plagiarism Detection with Wetrocloud
- REST API Implementation in Python for Model Deployment: Flask and FastAPI.
- Saving Your Machine Learning Model In Python: pickle.dump()
- Web Scraping with MS Excel and Python: Static Site Contents
- An exposé on Retrieval-Based ChatBot