Hi, I'm Adi 👋
Data and AI professional with a background in Computer Engineering from Mumbai University and a Master's in Data Science, Analytics, and Engineering from Arizona State University. My work sits at the intersection of analytics, machine learning, data engineering, and AI systems — from scalable pipelines to predictive models, dashboards, and intelligent AI-driven workflows.
Throughout my journey, I've had the opportunity to work, contribute, and collaborate with these organizations.
About me
I started in Computer Engineering at Mumbai University and went on to a Master's in Data Science, Analytics, and Engineering at Arizona State University.
My work sits at the intersection of analytics, machine learning, data engineering, and AI systems. I enjoy building end-to-end solutions — from transforming messy real-world data into scalable pipelines, to developing predictive models, interactive dashboards, and intelligent AI-driven workflows.
Across academic, research, and industry projects I've worked with large-scale datasets and distributed systems, predictive modeling and statistical analysis, ETL pipelines and cloud-based data workflows, business intelligence and visualization, and retrieval-augmented generation (RAG) and agentic AI systems.
I'm especially interested in problems where data drives measurable impact — operational efficiency, better decision-making, risk identification, or intelligent automation. Curious? Chat with an AI version of me about my background, projects, and experience.
How I work
Good decisions begin with evidence, not assumptions.
Finding patterns across messy, real-world data is where the fun begins.
Ship end-to-end solutions, learn from feedback, and iterate.
Every pipeline, model, or dashboard should move a real metric.
Portfolio
13 projects across autonomous AI agents, machine learning, analytics, and research — most with code or a write-up.
An autonomous AI co-worker that takes a goal, picks its own tools, evaluates its own work, and keeps going until the job is done.
Impact — Goal-driven agent loop with self-evaluation — it decides which tools to use and iterates without follow-up prompting.
AI agents collaboratively discover, research, and select high-potential stocks autonomously.
Impact — Hierarchical agent crew that researches, compares, and produces a stock recommendation end to end.
Autonomous AI sales outreach system.
Impact — Specialized agents draft, review, and dispatch outreach emails — a lightweight AI sales development pipeline.
A fraud detection pipeline combining XGBoost, LightGBM, NLP, and graph features, catching 82% of fraudulent transactions across 590K+ records.
Impact — 82% of fraudulent transactions detected across 590K+ records by fusing tabular, text, and graph signals.
An NLP and XGBoost-powered platform analyzing 123K+ job descriptions to extract in-demand skills, predict salaries, and cluster the job market.
Impact — 123K+ job descriptions parsed into skill demand signals, salary predictions, and market clusters.
Pre-tournament 2026 FIFA World Cup winner prediction using Elo ratings, XGBoost, and 10,000 Monte Carlo simulations of the full 48-team bracket.
Impact — 10,000 Monte Carlo simulations of the full 48-team bracket, driven by Elo ratings and an XGBoost match model.
A churn prediction model built with feature engineering and classification techniques — correctly flagging 84% of at-risk telecom customers before they leave.
Impact — 84% of at-risk customers correctly flagged before churn, paired with a retention strategy write-up.
A recall-optimized diabetes risk classifier using SMOTE and RFE, deployed as an interactive app for real-time risk scoring.
Impact — Recall-first modeling with SMOTE balancing and RFE feature selection, shipped as an interactive risk-scoring app.
A regression-based model estimating FIFA player market value from performance attributes and statistics.
Impact — Regression modeling over player performance attributes to explain and estimate market value.
County-level obesity prevalence estimation using CDC BRFSS survey data and ACS census demographics.
Impact — Combines CDC BRFSS survey responses with ACS census demographics to estimate prevalence at county level.
An interactive dashboard surfacing trends and key metrics across S&P 500 market data.
Impact — Turns raw market data into an interactive view of trends and key performance metrics.
A Google Trends analysis revealing how Grand Slam interest shifts by region — the US Open dominates in the US, while Roland Garros and Wimbledon lead across Europe.
Impact — Regional search-interest patterns show how each Grand Slam owns a different part of the world.
Published IEEE research comparing RNN and its variants for forecasting the Air Quality Index.
Impact — Peer-reviewed study published in March 2024 (IEEE Xplore).