Portfolio · LinkedIn · narasimharoyal31@gmail.com
Data analyst in Houston. I turn messy, multi-source data into numbers a business can act on, and I build the pipeline underneath when a project needs one.
| Area | Tools |
|---|---|
| Languages | SQL Python DAX |
| Analytics & BI | Power BI Excel pandas NumPy Tableau |
| Databases & Cloud | PostgreSQL SQL Server Snowflake BigQuery |
| ERP | SAP S/4HANA |
| Data Engineering | Airflow dbt Kafka Great Expectations Docker |
| Geospatial | PostGIS GeoPandas QGIS |
Question: How much do U.S. chip imports depend on a handful of countries?
Found: In some months a single country supplied up to 58% of U.S. memory-chip import value. Anyone can ask the data in plain English through a text-to-SQL assistant.
Kafka Airflow dbt PostgreSQL BigQuery Great Expectations Docker
Question: How often do a review's words disagree with its stars, and who is gaming the ratings?
Found: Across 44.2M Amazon reviews, the text and the rating disagree in about 1 in 5. A behaviour-based risk score flagged 1,037 reviewers for a manual look.
Python PostgreSQL Airflow VADER Power BI
Question: Where do two public maps of Houston's power grid disagree about where things are?
Found: 474 findings across central Houston, ranked by FEMA flood risk so the riskiest get checked first.
PostGIS GeoPandas QGIS Python
Question: Does the daily risk number a gas trading desk relies on hold up in a winter storm?
Found: Before the January 2024 cold snap it said 15%. The real move was 319%. Tested on 29 years of public price data.
Python PostgreSQL scikit-learn statsmodels React
Question: Which battery storage site is about to run short, and is production keeping pace?
Built: Daily inventory simulated in Python and forced to add back up to Tesla's real quarterly figures from its SEC filings, on a star schema with three Power BI pages.
Python PostgreSQL Power BI (DAX) SQL


