I frame the decision
before I fit the model.

I'm Rajan — an industrial engineer by training, now working as a data scientist across data science, analytics, and decision science. Most teams aren't short on data; they're short on the decision it should drive. I build the models, analysis, and optimization that close that gap.

PIPELINE // ACTIVE
18ms
Measured Outcome↑ +34.8% ROI
+34.8%Operational efficiency & predictive accuracy
Optimization TrajectoryBaselineOptimized ML
Annualized ROI+$240K
ROC-AUC0.94
Precision98.2%
query.sql1.2M rows
PostgreSQL · Feature Extraction
model.fit() + OR-ToolsOptimal
XGBoost & Mathematical Solver
Automated Action+34.8% Gain
Operational Decision Dispatch

Background

My background in industrial engineering deeply informs how I approach data. Industrial engineering is the discipline of optimizing systems under constraints — which is exactly what decision science is, just with a model in the middle.

I moved into data science because I kept seeing the gap between a dashboard full of numbers and the decision those numbers were supposed to inform. I like closing that gap.

What I do

I turn business data into models, analysis, and decisions — across data science, analytics, and decision science.

Predictive Modeling

Churn, LTV, demand, and lead-scoring models — from feature engineering to deployment.

Optimization

Pricing, allocation, and routing framed as decisions and solved as models.

Tech Stack
PythonSQLscikit-learnpandasPostgreSQLTableau

Experience

Where I've turned business data into decisions.

ING Consulting

Data Analyst
Jan 2026 — Present · 8 mosFull-time
Kathmandu, Bāgmatī, Nepal · On-site

Delivered enterprise automation, predictive analytics, and data pipeline solutions across 13 total entities—including 9 educational institutions (partnered with London Metropolitan, Wolverhampton, and Pokhara Universities) and 4 corporate subsidiaries (Innovate Tech, Vairav Tech, ING Consulting, ING Skills).

  • System Automation: built and deployed an automated student feedback system handling 10,000+ submissions, cutting response time from 5s to under 50ms (100x faster) across 10 institutions via a multi-layer caching architecture.
  • Demand Forecasting: built demand forecasting models for admissions and business operations to support capacity planning.
  • Reporting Automation: automated cross-college reporting pipelines for enrollment and satisfaction KPIs, eliminating 8+ hours of manual reporting per week.
  • Process Automation: built a PAT (Personal Academic Tutor) management system automating supervisor routing and feedback collection across all campuses, eliminating manual coordination.
  • Budget Automation: built a budget automation system that consolidates monthly expense data across 10 colleges and flags budget anomalies.
  • HR Payroll Repo: built a repository for the complete calculation and audit of staff salary, including a visual dashboard.

rajanbuilds

Founder & Decision Automation Consultant
Oct 2024 — Present · 2 yrsSelf-employed
Kathmandu, Bāgmatī, Nepal · Remote

These are the types of solutions I build for businesses:

  • Predictive Analytics: built a customer churn and revenue forecasting system to support proactive business decisions.
  • Optimization: developed a workforce scheduling model to optimize resource allocation while minimizing operational costs.
  • Business Analytics: built an executive KPI dashboard for sales, finance, and operational performance monitoring.
  • Decision Support Systems: developed a pricing decision engine using demand patterns, costs, and business constraints.
  • Workflow Automation: automated inventory monitoring and replenishment workflows with real-time alerts and reporting.

How I work

Project Engagements

For a defined question or decision. We frame the problem and success metric upfront, agree on deliverables, and I build the analysis or model. No open-ended hourly surprises.

  • Problem framing & data audit
  • Model / analysis build & validation
  • Documentation & handoff

Analytics Partner

For ongoing analysis and models that need to keep earning their keep. Consider me your fractional data scientist — monitoring, retraining, and answering the next question as it comes.

  • Model monitoring & retraining
  • New analyses & dashboards
  • Priority support

Full breakdown on the pricing page.