SELECT insight FROM data; -- hello, I'm

AshRana

Automation-Focused Data Analyst | AI Agents & Business Intelligence

I dig into business problems, surface the trends hiding in the data, and build AI-powered systems that automate the manual work.

View my workGet in touch →
SELECT * FROM services;

What I do

Three ways I make data useful — from the first query to the finished decision.

🤖

AI & Automation

LLM-powered tools and agents that take repetitive work off people's plates, from a pricing audit pipeline that took a 3.5-hour manual cycle down to about 12 minutes of review, to an AI agent that queries a database in plain English.

🔍

Data Analysis & Insight

SQL, Python, and Excel across millions of rows: cleaning, joining, and interrogating real datasets until the actual trend shows up. Conclusions that hold up to scrutiny, caveats and all.

📊

Business Intelligence

Interactive Tableau and Power BI dashboards that turn analysis into something a decision-maker can act on in seconds. Built at scale, including a catalog of 475K+ SKUs reviewed weekly by leadership.

SELECT role, org FROM career ORDER BY year DESC;

The journey so far

In progress · Dec 2026

M.S. Computer Science (AI & Machine Learning)

Western Governors University

Graduate coursework in machine learning, deep learning, NLP, and AI system design. Building the engineering rigor beneath the analytical work.

Jun 2025 – Jan 2026

Product Data Analyst

United Aqua Group

Rebuilt the quarterly vendor rebate model in SQL, joining sources that each sat at a different grain, and recovered rebate the old process had missed. Built a pricing pipeline that took a vendor audit from 3.5 hours to about 12 minutes, and led the 475K-SKU catalog migration into NetSuite. Ran the Tableau and Power BI reporting leadership used weekly, plus a 3-person team on LLM-driven catalog enrichment.

Mar 2025 – Jun 2025

Business Support Specialist

United Aqua Group

Cleaned and validated vendor pricing files of up to 5,000+ SKUs for NetSuite and BigCommerce, reducing pricing mismatches by roughly 30% and upload rejections by 15%+.

Aug 2023 – Feb 2025

Business Data Analyst

Healthy Delights · Remote

Delivered Power BI dashboards, Power Apps, and Power Automate flows used across the company. Built the seasonal demand and inventory models that cut stockouts on top sellers by roughly two-thirds, and owned reporting on revenue, customer acquisition cost, and repeat purchase behavior that informed executive decisions, and managed 3 interns.

Oct 2022 – Jul 2023

Junior Analyst

Healthy Delights · Remote

Centralized purchase, stock, and vendor cost data for 50+ items, enabling 3x faster monthly reporting. Audited and cleaned 6+ months of order history, cutting pricing errors by roughly 25%.

Graduated

B.S. Hospitality Management (Entertainment Business)

University of Nevada, Las Vegas

Where I learned how businesses actually make money: revenue models, operations, and customer behavior. The analytical instinct came later. The commercial one started here.

SELECT * FROM projects WHERE impact = true;

Featured projects

Real dashboards, real pipelines. Code on GitHub where it can be shared.

ai-data-analyst-agent.app
AI Data Analyst Agent dashboard

AI Data Analyst Agent

A deployed, autonomous agent that answers plain-English business questions by inspecting a database's schema, writing and running its own SQL, recovering from its own query errors, and explaining the results through a live web app with a transparent decision log. Built with read-only guardrails and safety limits. The natural next step from the campaign analysis: first I did the analysis by hand, then I built the system that does it on demand. Try it live at ashranawork-ai-agent.streamlit.app.

PythonGemini APIAI AgentsStreamlit
market-pulse.app
Market Pulse dashboard

Market Pulse

An event-driven ELT pipeline producing daily hospitality demand signals for Las Vegas and Los Angeles from live flight and weather data. Lambda on an EventBridge cron and GitHub Actions land raw data in a partition-projected S3 lake, dbt transforms it through staging and mart layers behind 12 data-quality tests that fail the build on violation, Athena queries it, and SNS alerts on high demand pressure. The transform layer is then implemented a second time in PySpark writing Delta Lake tables on Databricks, with a validation suite of 7 assertions proving both engines produce identical output. Runs unattended on a schedule for roughly $1/month.

PythonAWSdbtPySparkDelta LakeDatabricksGitHub Actions
pricing-discrepancy-detection-tool.app
Pricing Discrepancy Detection Tool dashboard

Pricing Discrepancy Detection Tool

An automated auditing tool that compares vendor invoices against agreed contract prices, flags overcharges, exports a clean exception report, and drafts vendor correction emails with an LLM. I built the production version of this at United Aqua Group; this is an independent rebuild on synthetic data so the code can be public. On that data it flagged $5,810 in overcharges, the kind of manual, error-prone check that quietly leaks money at real companies.

PythonPandasOpenAI API
retail-marketing-campaign-analysis.app
Retail Marketing Campaign Analysis dashboard

Retail Marketing Campaign Analysis

An end-to-end analysis of 2.6 million retail transactions across 2,500 households, asking whether marketing campaigns actually drive incremental spend. The work separates genuine campaign response from selection bias and lands on a concrete targeting recommendation, presented in an interactive Tableau dashboard.

SQLSQLiteTableau
More on GitHub →
INSERT INTO inbox (message) VALUES (yours);

Let's build something with your data

I work where analysis meets automation: understanding how a business runs, finding what the data says, and building the system that does the manual work on its own. If you've got a business problem buried in data, or a manual process eating hours it shouldn't, let's build something to solve it!

Email meLinkedIn ↗