MELVIN/m3lv1n
DataSoftwareBlogAboutContact
home/data projects

Data
Projects

Four questions I couldn't stop poking at — labour markets, viral video, supermarket price wars and a leaking sales funnel. Each one is written as a data story: a finding first, the working underneath.

4
Case studies
0
Rows analysed
/ 01ABS · Azure SQL · Power BI
80%
of working men hold a full-time job — for women it's 57%. Australia's part-time economy is overwhelmingly female.

Australian Labour Market Dashboard

An end-to-end pipeline on live ABS data — Python ingests it, Azure SQL models it through a staging→mart layer, and a 4-page Power BI report visualises it. The report is generated as code, not clicked together.

ABS APIAzure SQLPower BIDAXPython
Read the case study ↗
/ 02API · pandas · ML
38 hrs
The average video's entire life on the Trending page lasts under two days — then it vanishes.

YouTube Trending Analytics

Forensics on 40,000 trending videos across 10 regions: what actually predicts a spot on the board, how long it survives, and which signals are pure noise.

YouTube APIpandasscikit-learnPlotly
Read the case study ↗
/ 03Scraping · dbt · Postgres
$1,240
The yearly gap on an identical basket of staples — but the cheaper store flips week to week. Specials decide it, not shelf price.

Woolworths vs Coles Price Analytics

Six months of daily price scrapes on 1,000 staple products. Who's genuinely cheaper, when each store discounts, and how much loyalty actually costs you.

Web ScrapingPostgreSQLdbtPython
Read the case study ↗
/ 04Snowflake · dbt · Looker
3.1×
Deals with a demo booked in week one closed at triple the rate — regardless of company size or deal value.

SaaS Sales & Revenue Pipeline

A full funnel teardown of 4,200 opportunities — where revenue leaks, where deals stall, and the two behaviours that compound into closed-won.

SQLSnowflakedbtLooker
Read the case study ↗

Also a builder

Want to see what I ship in code?

View Software Projects ↗