Satyam Sharma

About Me

Hi, I'm Satyam Sharma

I'm a second-year BA Economics (Hons) student at Maharshi Dayanand University, Rohtak (2025–2029) with a minor in Mathematics, building a career in data analysis. My core tools are Power BI, MySQL and Excel, and I'm learning Python (pandas).

From July to September 2026 I worked as a Data Operations Trainee at the Municipal Corporation of Gurugram: keeping records for 95 solid-waste vehicles, cleaning GPS data in Excel and sending daily reports to senior officials. Alongside that I've built four end-to-end analytics projects, listed below.

One habit I care about: checking my numbers before I present them. When I re-audited my own logistics dashboard against the raw data, three data-entry errors had changed my main conclusion. The Data Audit section shows what I found.

I'm looking for data analyst, MIS or reporting roles and internships, and I take on small freelance dashboard projects.

📍 Faridabad, Haryana 🎓 BA Economics (Hons) 💼 Internships & entry-level roles 🤝 Freelance projects

Experience

Data Operations Trainee

Municipal Corporation of Gurugram · Full-time, on-site

Jul 2026 – Sep 2026

Maintained records of 95 solid-waste-management (SWM) vehicles (53 in Zone 6, 42 in Zone 8) on the SWM Portal: added, removed and updated GPS IDs and mapped drivers, helpers and supervisors.
Monitored daily vehicle movement on the GPS tracking website and downloaded data from the GPS site and the SWM Portal.
Cleaned and consolidated data in Excel (Remove Duplicates, SUMIF, COUNT, Pivot Tables), flagged data errors and prepared daily reports.
Uploaded reports to the official reporting website and emailed the daily GPS report to senior MCG officials by 5:30 PM.
Attended meetings at the Gurugram Head Office and explained data to government officials to resolve vehicle penalties.

ExcelPivot TablesMIS ReportingData Cleaning

Featured Projects

Self-initiated portfolio projects, each from raw data to dashboard.

Logistics Performance and Cost Dashboard — page 1

Logistics Performance and Cost Dashboard — page 2

Google SheetsApps ScriptPower BIDAX

Logistics Performance & Cost Dashboard (3PL)

A Google Sheets → Power BI reporting flow for 2,459 shipments, 7 carriers and 10 warehouses (Jun 2023 – May 2024), with an Apps Script trigger that emails a daily "data entry complete" alert.

147 lost shipments (5.98%) worth $34.6K in shipping cost; 287 delayed (11.67%).
USPS is the cheapest carrier per shipment (about $204) but loses the most: 8.3% vs 3.8% for UPS.
Boston and SF ship about 10% slower on average, mostly because their routes are longer.

Practice dataset · profit is modelled at a flat 30% markup

The screenshot above predates my data audit. Corrected figures are in the Data Audit section.

View Project →

Texas Real Estate Market Dashboard — page 1

Power BIPower QueryDAXAzure Maps

Texas Real Estate Market Dashboard

No ready-made CSV: I extracted 478 property listings from a real estate site's JSON network responses (browser Network tab), cleaned them with VS Code and Power Query, then built an interactive dashboard with city, beds, baths and property-type slicers, price-category analysis, custom DAX measures and an Azure Maps listing-density map.

478 listings · JSON extraction · custom DAX measures

View Project →

Credit Card Fraud Dashboard — second page

MySQLPower BIDAX

Credit Card Fraud Analysis

100,000 sample card transactions loaded into MySQL, analysed with 4 SQL KPI queries (total fraud loss, top cities, fraud by transaction type, peak hours) and presented in a Power BI dashboard.

1,000 of 100,000 transactions (1.00%) are flagged as fraud, worth $2.52M.
Fraud was spread almost evenly: refunds 50.7% vs purchases 49.3% of cases; no hour, city or weekday stood out statistically.

100,000 sample transactions · 4 SQL KPI queries · MySQL → Power BI

View Project → ▶ Live Dashboard

Samsung Global Sales — Executive Overview

Samsung Global Sales — Customer and Product

Samsung Global Sales — Return and Performance

MySQLPower BIDAX

Samsung Global Sales Analytics (2021–2024)

A MySQL database (schema, import, indexes) of 15,500 transactions across 52 countries, with 7 analytical query blocks on category, region, channel, returns and year-over-year growth, and a 3-page Power BI dashboard (Executive Overview, Customer + Product, Return + Performance).

Galaxy Z has the highest return rate: 10.87% of orders (1,058 orders).
Returned orders total $2.19M, 9.6% of revenue (all channels).

15,500 transactions · 28 columns · 52 countries · 7 query blocks

View Project → ▶ Live Dashboard

Data Audit: I Checked My Own Dashboard

Before publishing this portfolio I re-computed every headline number in my Logistics dashboard from the raw 2,459-row dataset (with AI assistance) and compared it with what the dashboard showed. Three data-entry errors had changed my main conclusion.

Metric	My dashboard said	The raw data says
Total shipments	“2K” (card rounding)	2,459 shipments.
Most expensive carrier	UPS, $235 per shipment	UPS is $212 once 3 extreme-cost rows are excluded ($6,562, $6,229 and $1,598; the next-highest shipment in the dataset costs $783). FedEx ($228) and DHL ($225) cost more.
Costliest warehouses	Houston ($242) and Los Angeles ($233)	The same 3 rows inflated both. Cleaned, Seattle ($226) is highest; Houston is $218 and LA is $204.
“Shift UPS volume to USPS to cut cost 10–15%”	Cheaper carrier, so move volume	USPS saves $7.85 per shipment (3.7%) but loses 8.3% of shipments vs 3.8% for UPS. Counting a lost shipment's cost as lost, the extra expected loss (about $8.85 per shipment) cancels the saving. Moving all UPS volume would change total cost by about 0.5%.
Total profit	$161K	$128K on delivered and delayed shipments. $32K had been counted on lost, returned and in-transit shipments. In this dataset profit is a flat 30% markup on cost, so it is modelled, not observed.
On-time delivery	70%	75.6% once In Transit shipments are excluded. The 70% treated shipments that aren't late yet as late.
Boston and SF “10% slower”	About 10% slower than Denver and NYC	True on raw averages (4.77 and 4.64 vs 4.29 and 4.30 days), but per 1,000 miles the gap is only about 3%. These warehouses ship longer routes.

Root causes

4 Revenue and 2 Profit values were stored as text with thousands separators (e.g. “8,097.86”), there was no outlier check, and I hadn't written down what each KPI includes.

What I took from it

Re-compute headline KPIs from the raw data, look for outliers and text-typed numbers, and state what a metric includes (like profit on lost shipments) before turning it into a recommendation.

Skills & Tools

Power BI & Excel

Power BI: Power Query, DAX (basic), slicers, bookmarks, Azure Maps. Excel: Pivot Tables, SUMIF, COUNT, VLOOKUP, XLOOKUP, Remove Duplicates, MIS-style daily reports.

SQL

MySQL: SELECT, WHERE, GROUP BY, aggregate functions, data import and cleaning. HackerRank SQL (Intermediate) certified.

Python

pandas basics for cleaning and checking data. Currently learning, and using it to audit my own projects.

Data Handling

Data cleaning, KPI dashboards, JSON extraction (browser Network tab, Power Query), Google Sheets, Apps Script time-driven triggers.

Tools

MySQL Workbench, Power BI Desktop, VS Code, GitHub. I use Claude and Gemini to speed up SQL, DAX and Excel formulas, and I check their output.

Reporting & Communication

Daily MIS reports delivered on time to senior officials, and data explained in meetings with government officials. Accurate entry, clear written updates and working with a team.