Data Analyst · BI Analyst
I turn messy data into decisions people actually act on. Whether it's a Power BI dashboard, a SQL pipeline, or a geospatial analysis, I care about making the output clear, useful, and impossible to ignore.
Fair warning: AI had a hand in this. But the data work, the insights, and the late nights untangling messy datasets? That was all me.
about
I'm drawn to the moment when a messy spreadsheet or a raw data dump becomes something a stakeholder can actually use to make a decision. That's what got me into data analytics, and it's still what gets me excited about the work.
My background spans SQL querying and data wrangling, building Power BI dashboards for real operational teams, GIS-based geospatial analysis, and writing PySpark pipelines that process data at scale. I care about clean outputs, clear visuals, and making sure the insight actually lands. I also paint, doing detailed intricate work that needs the same patience as getting a dashboard right.
tech stack
Analyzed 3,611 county-year records merging OpenFEMA disaster assistance data with Census ACS socioeconomic indicators. Performed full EDA, feature engineering, and multi-model comparison to surface a key policy finding: applicant volume drives FEMA disbursements far more than poverty rate, challenging how relief is prioritized. Built a live Streamlit dashboard to make the findings accessible to non-technical audiences.
Built an end-to-end analytics pipeline processing 50K+ IoT network records using PySpark. Designed Power BI and Kibana dashboards that translate complex threat patterns into clear, actionable visuals for non-technical stakeholders.
Built for a real manager's workflow: an ML pipeline using sentence-transformers and Logistic Regression to automatically sort P&C insurance PDFs into 14 categories. Achieved 98.6% accuracy, cutting manual document triage time significantly.
Automated data quality auditing engine that scores datasets for missingness, outliers, and feature leakage. Produces a trust score (0 to 100) with risk labels so analysts can catch upstream problems before they corrupt reporting downstream.
End-to-end analytics project on 397,000+ BPD arrest records. SQL analysis in Python, data cleaning and wrangling, then a published Tableau dashboard covering arrest trends from 2015 to 2026, top offense types, district hotspots, time-of-day patterns, and demographic breakdowns.
View Live Dashboard ↗Data Analyst Intern
UMBC Transportation Services · Baltimore, MD
Descriptive, predictive, and prescriptive analytics for campus transit operations. Building dashboards for real-time monitoring, GIS-based geospatial route analysis, and defining KPIs to track on-time performance and service coverage.
Customer Support Executive – Analytics
PetBacker · Ahmedabad, India
Analyzed customer interaction data to identify response bottlenecks, driving a 20% operational efficiency improvement. Built Power BI dashboards tracking user behavior and engagement metrics used in cross-functional decision-making.
Data Analysis & Machine Learning Intern
(n)Code Solutions · Gandhinagar, India
EdTech analytics internship where I built a predictive model achieving 92.9% accuracy on student performance data. Processed 50,000+ IoT records via PySpark ETL pipeline and presented findings to technical and non-technical stakeholders.
Master of Professional Studies in Data Science · GPA 3.822
University of Maryland, Baltimore County (UMBC) · Baltimore, MD
Capstone: FEMA Disaster Resource Allocation. I analyzed 3,611 county-year records, built a live Streamlit dashboard, and surfaced a key policy finding on how disaster relief actually gets distributed. Coursework in data analytics, big data systems, SQL, visualization, and applied ML. OPT authorized.
B.Tech in Information Technology · GPA 3.36
Institute of Advanced Research (IAR) · Gandhinagar, India
Undergraduate degree in Information Technology covering programming, data structures, algorithms, and software engineering fundamentals.
Looking for full-time Data Analyst and BI Analyst roles where the work has real impact. I love turning raw data into dashboards and decisions that stakeholders actually use. Open to remote, hybrid, or on-site in the U.S. OPT authorized, no sponsorship required to start.