Turning data into
decisions
I'm Rajan — a data scientist and industrial engineer. I help operations and product teams optimize workflows, automate predictive pipelines, and turn complex data sets into actionable ROI.
Python
SQL
Machine Learning
Optimization
PIPELINE // ACTIVE
18ms
Optimization TrajectoryBaselineOptimized ML
Annualized ROI+$240K
ROC-AUC0.94
Precision98.2%
query.sql1.2M rows
PostgreSQL · Feature Extractionmodel.fit() + OR-ToolsOptimal
XGBoost & Mathematical SolverAutomated Action+34.8% Gain
Operational Decision DispatchPythonSQLMachine LearningOptimizationpandasscikit-learnPostgreSQLTableau / Power BIPythonSQLMachine LearningOptimizationpandasscikit-learnPostgreSQLTableau / Power BI
Core Competencies
Data Science
Predictive and causal models — churn, LTV, demand forecasting, and lead scoring. From feature engineering in Python to models that ship, get monitored, and are retrained on new data.
Data Analytics
SQL-first analysis and dashboards that answer the actual question. Turning raw warehouse tables into clean metrics, cohorts, and the insight leadership acts on.
Decision Science
Optimization and experimentation — pricing, budget allocation, routing, and A/B testing. Framing the decision first, then building the model that makes it.
Frequently asked questions
- Who is Rajan Shrestha?
- Rajan Shrestha is a data scientist and industrial engineer working across data science, data analytics, and decision science. He uses Python, SQL, machine learning, and optimization to turn business data into models, analysis, and decisions teams can act on.
- What does Rajan build?
- Predictive and causal models (churn, LTV, demand forecasting, lead scoring), SQL analytics and dashboards, and decision-science systems — pricing optimization, marketing-budget allocation, and route optimization. Each project is built to ship and be maintained, not left as a one-off notebook.
- What tools and technologies does he use?
- Python (pandas, scikit-learn, XGBoost, statsmodels) for modeling, SQL and PostgreSQL for analytics, optimization libraries like OR-Tools and PuLP/SciPy for decision problems, and Tableau / Power BI for visualization. dbt for the metric layer.
- What is the difference between data science, analytics, and decision science?
- Data analytics explains and measures what happened; data science predicts what will happen; decision science optimizes what to do about it. Rajan works across all three so the answer doesn’t stop at a chart — it ends in a decision.
- What kinds of problems does he work on?
- Pricing, marketing spend, logistics, revenue and operations — anywhere a decision can be framed as a model. If the problem looks like one of the projects, it is probably a fit.