WorkResearchRésuméAboutContact

Research

Real methods. Real data. My research is about what happens when you apply ML to how people actually behave online. The unusual part: I am the primary data source.

IEEE ForthcomingPrimary AuthorIn Progress

Measuring Influencer ROI: Detecting Fake Engagement and Quantifying Authentic Reach on Instagram

This research addresses a structural gap in digital marketing analytics: how do brands accurately measure return on influencer investment when engagement metrics can be artificially inflated? The study uses real Instagram data from my own account @aneela_veldi as the primary dataset — 4,823 posts over three years, which is rare. Most influencer research uses scraped or synthetic data. This was mine, verified, and real.

We built models combining anomaly detection, NLP comment analysis, and time-series engagement modeling to separate real reach from bots. The output is a scoring method brands can understand and use when they're deciding where their influencer budget actually goes.

PythonNLPAnomaly DetectionTime SeriesPandasMatplotlibSeaborn
Total impressions analyzed8M+
Avg gap: inflated vs. authentic23%
Bot detection F1-score0.91
Viral posts: Wed–Fri, 6–9 pm73%

IEEE Conference on Data Science and Advanced Analytics (DSAA) / International Journal of Information Management

First-party, longitudinal Instagram data from @aneela_veldi. 4,823 posts · 36 months · verified engagement.

ISB Research · 2022–2023

Emerging Market Consumer Behavior Analysis

Research Associate at the Indian School of Business, Asia's #1 globally ranked business school. Studied how people make buying decisions in South Asian markets. Managed panel datasets of 100K+ records, ran econometric models in Stata and R, and worked with faculty on peer-reviewed publications.

StataRPythonEconometricsPanel Data
DePaul · 2024–Present

Enrollment Prediction & Admissions Analytics

Applied predictive modeling to law school admissions data at DePaul College of Law. Built enrollment forecasting models using regression and time-series techniques, and automated reporting infrastructure that replaced 20+ hours of weekly manual analysis. Findings presented directly to the Dean's office to guide which students to recruit and when to push harder on yield.

PythonTableauSQLForecastingRegression

The Education Behind the Analysis

2024 – 2026

DePaul University

M.S. Business Analytics

Advanced data modeling, ML, predictive analytics, strategic business intelligence

In Progress
2020 – 2022

Osmania University

MBA — Finance & Operations

Corporate strategy, financial analysis, operations management, econometrics

Completed
2017 – 2020

Osmania University

Bachelor of Commerce

Business fundamentals, accounting, quantitative methods, economics

Completed