Pranav Kaja, Data & Analytics Engineer
Data & Analytics Engineer. I build pipelines, models, and dashboards that survive contact with real data.
Origin File01. ABOUT
ORIGIN STORY
I started in computer science and moved into business analytics, and finished my MS at UMass Amherst to bridge writing code and driving strategy. I care less about the model that looks perfect on a slide, and more about the one that survives contact with real data.
APPROACH
Pragmatism over perfection, but pragmatism doesn't mean cutting corners. A model that ships and gets monitored beats a perfect one stuck in a notebook. I build for the whole lifecycle and leave an audit trail the next person can actually read.
Ops Log02. EXPERIENCE
UMASS AMHERST DINING
Student Supervisor
Automated 37-location ETL pipeline, cutting manual reporting from 2 hrs/day to 12 min/day. Led a team of student employees and implemented data-driven inventory forecasting.
ACMEGRADE
Data Science Intern
Trained Random Forest and Gradient Boosting on a 100,000-record client portfolio, lifting conversion-prediction accuracy 15% over baseline.
Mission History03. PROJECTS
Invitovate
Postgres + RLS / Vanilla JS
A calm, multi-tenant outreach pipeline. Draft in your voice, send only what you approve.
Overheard
BigQuery / Vertex AI (Gemini)
Finds and ranks the customer questions brands ignore on their own social posts. 17 brands, live.
CPK Production Planner
In-Browser Tool / JS
Order report in, make lists and pick sheets out, across 36 locations.
Fraud Triage
XGBoost / Gemini
0.978 AUC; caught fraud the base model missed.
Corporate Analytics
Power BI / SQL
30% faster executive reviews via custom dashboards.
Hospital Readmission
Python / Gradient Boosting
90% recall on readmissions via threshold tuning.
Loadout 04. SKILLS
[+]CORE TECHNICAL STACK
DATA ENGINEERING
SQL, ETL/ELT, Python, Spark & Hadoop, Data Warehousing
ANALYTICS & BI
Power BI, Tableau, Power Query, KPI Design, Financial Modeling
MACHINE LEARNING
XGBoost, Random Forest, Logistic Regression, SMOTE
PROGRAMMING
Python, SQL, R, VBA, REST APIs, LLM Integration
STRATEGY & OPS
Demand Forecasting, Supply Chain Optimization, Capacity Planning
CERTIFICATIONS
Data Analysis with Python
IBM
Intro to Big Data with Spark & Hadoop
IBM
Supervised Learning with scikit-learn
DATACAMP
Ensemble Methods in Python
DATACAMP
Working with Hugging Face
DATACAMP
Database Management Essentials
UC COLORADO
Side Quests05. INTERESTS
BEYOND THE SCREEN
A glimpse into what drives me when I'm not building systems.
ACCESS INTEREST_LOGS EXPLORE MY INTERESTS