Chris Perry

Make intervals easier to read with SQL

FULL JOIN = membership in both directions. Employees without a department and departments without employees show up in one query. Careful: COUNT(*) counts the NULL rows too.

Millions of rows. A handful of insights. One decision. Great analysts don’t just analyze data. They reduce complexity, identify what matters, and turn information into action. The goal isn’t more information. It’s better decisions. #dataanalytics #businessintelligence #dataanalysis

SQL gets much easier when you understand how the core clauses work together. SELECT → FROM → WHERE → GROUP BY → HAVING → ORDER BY Each clause has a distinct job. The real skill is knowing when to use each one, and how they combine into a complete query.

Trend tells you where the data is going. Seasonality tells you what keeps repeating. A time series can have both at the same time. Separating them is key to understanding what’s actually changing. #analytics #datascience #forecasting

Your text-to-SQL chat history is more than a log. It’s a record of how users actually query your data. Repeated columns, filters, and date ranges can reveal opportunities for better indexing and partitioning. This is just one example using Salesforce-themed chat logs which are simplified in the graphic. Don't get too caught up in the phase 1 and phase 2 recommendations. Again, they are just examples and would vary based on the data and how users interact with it. #database #ai #sql

Two datasets can have the same mean but look completely different. Standard deviation tells you what the mean can’t: how spread out the data is. #statistics #datascience #analytics

Data leakage happens when a model learns from information it wouldn’t have at prediction time. Great test results. Poor real-world performance. Always train with what the model would actually know. #MachineLearning #DataScience #AI

Cleaner pandas code isn’t about writing less. It’s about making transformations easier to read, reuse, test, and trust. Method chaining + .pipe() can turn scattered steps into a real pipeline. #pandas #python #dataengineering

The future of analytics may not be better dashboards. It may be conversations. Anthropic automated ~95% of analytics questions by treating analytics as a knowledge problem, not a SQL problem. #dataanalytics #genai #ai

LISTAGG is one of the most practical Oracle SQL functions for reporting. Turn repeated rows into clean, ordered summaries directly in SQL. Great for exports, dashboards, and audit reports. #OracleSQL #SQL #DataEngineering

Dashboards should guide the next click. Tableau action filters turn one view into a control for another. Less clutter. More context. Faster answers. #datafam #tableau #dataanalytics

List comprehensions look confusing… until you realize they follow one simple pattern: what → for → in → if Once that clicks, they become one of Python’s most readable features. #python #datascience #coding

Many pandas nested data problems come from this: 1 row contains multiple values. DataFrame.explode() fixes that by turning: 1 row with many values to many rows with 1 value each Perfect for APIs, JSON, tags, surveys, and more. Small method. Huge impact. #python #datascience

Competitive Fortnite and database optimization have more in common than people think. Bad rotation = full table scan Efficient tunneling = proper indexing Storm surge = system bottlenecks Piece control = query optimization Different game. Same mindset. #Fortnite #DataEngineering

One-to-many joins can replicate rows. If one ACCOUNT has 3 OPPORTUNITY records, the ACCOUNT row appears 3 times after the join. That can inflate counts and aggregates. Know the relationship. Understand the grain. Write better SQL. #SQL #DataAnalytics