Experience
Industry · 2023 — 2026
Forecasts thatshipped.
From cruise-line revenue to enterprise PMO analytics to regulatory RAG — get the data to the right grain, then build something stakeholders can interrogate.
Career timeline
Education
Kumaraguru College
BTech IT
2020 – 2024
Columbia University
MS Data Science
2025 – 2026
Industry
Saama
DS Intern
2023 – 2023
E2Open
DS Analyst
2024 – 2025
Holland America
DS Intern
2026 – 2026
Part-time
People+ai
AI/ML Engineer
2024 – 2025
Research
Research projects
IRI · Columbia Climate School · NASA GISS · Mailman · CUIMC Radiology · Fulton Ring
2026 – 2026
Kumaraguru College
BTech IT
Columbia University
MS Data Science
Saama
DS Intern
E2Open
DS Analyst
Holland America
DS Intern
People+ai
AI/ML Engineer
Research projects
IRI · Columbia Climate School · NASA GISS · Mailman · CUIMC Radiology · Fulton Ring
Holland America Line
Business Strategy & Performance Analytics Intern
Jun 2026 — Aug 2026 · Miami, FL · Carnival Corporation
- Built and shipped a voyage-level revenue forecasting system (Python, Snowflake) at 3–4% WAPE against closed actuals — an explainable glass-box GAM adopted by multiple teams.
- Built a Snowflake Cortex–powered natural-language assistant so stakeholders could query the forecast directly.
- Causal revenue decomposition for HIA packages using Double ML (EconML).
- Pricing and price-elasticity analysis for Asia shore excursions.
3–4%
WAPE vs. actuals
E2Open
Data Science Analyst — PMO Division
Aug 2024 — Aug 2025 · Bangalore, India
- Defined the KPI framework (SLA attainment, cycle time, project-health metrics) that senior directors used to prioritize continuous-improvement initiatives.
- Built predictive models and executive dashboards used across departments, owning both the SQL and BI layers — cutting reporting turnaround from ~3 days to a few hours by replacing a manual process.
3d → hrs
Reporting turnaround
People+ai
AI/ML Engineer (part-time)
Aug 2024 — Mar 2025 · Bangalore, India
- Built NLP extraction and knowledge-graph pipelines over regulatory documents, and FindMyCircular — a Self-RAG + knowledge-graph chatbot returning cited answers.
RAG
Cited answers
Saama Technologies
Data Science Intern
Feb 2023 — Dec 2023 · Pune, India
- Benchmarked stock-yield forecasting models; Gradient Boosting won at 93% accuracy versus 87% for SVM.
93%
GBM accuracy
Teaching
In the classroom.
COMS 4776 Neural Networks & Deep Learning
Teaching Assistant
Prof. Richard Zemel · Columbia University
GR5015 Data Analysis for the Social Sciences
Teaching Assistant II
Quantitative Methods in the Social Sciences · Columbia QMSS
Computer Systems
Teaching Assistant
Prof. Waqar Aqeel · Columbia University
Leadership & community
Beyond the desk.
Vice President
Young Indians chapter
Member
Decoded Futures (Tech:NYC)
AI onboarding support
GIIN