Data Analyst (SQL + Excel + BI)

Real portfolio pieces built during the course, not a certificate for its own sake. Each one is work you can show.

• Cleaned dataset and cleaning logThe Fernway order data after deduplication, timezone conversion and category mapping, plus a log recording every issue found, how you found it, the rows affected and what you decided to do about it.
• Data dictionaryOne row per column of the cleaned dataset: type, business meaning, how often it is null, and the gotcha that would trip up the next person to use it.
• Documented SQL query setEight commented queries answering real business questions, each with a header stating the question, source, grain, definition, period and caveats. Covers profiling, deduplication, timezone handling, window functions and cohorts.
• Dashboard, and the spec behind itA one page specification, audience, the decision it supports, refresh, grain, metric definitions, filters, what is deliberately out of scope and one stated question per page, plus the dashboard built from it: a star shaped model, four measures, and one figure checked against SQL before anything else was drawn.
• One page insights memoHeadline, recommendation, quantified evidence, what it is worth with the assumption stated, caveats, and method. The artifact that proves you can commit to a conclusion.
• Agreed analysis briefThe vague request turned into something answerable and agreed before any SQL is written: the decision it feeds and when, the comparison hiding behind the word being used, the measure and its definition, what changed in the period, what is deliberately out of scope, and the three sentence brief sent back for a yes.