Matt Mike

Matt MikeOct 3, 2026

Apple is hiring Data Analysts! Check these out while they're still open: 1. Business Insights Analyst, Apple → https://lnkd.in/eUxB8xWA 2. Data Analyst, Loyalty Strategy, Starbucks → https://lnkd.in/e75m68Yf 3. Data & Insights Analyst, Visa → https://lnkd.in/ewVR85Ss 4. Analytics Specialist, Delta Air Lines → https://lnkd.in/exEAAm6R 5. Senior Analyst, Data Analytics, American Express → https://lnkd.in/epqMKUKx 6. Technical Planning Analytics Analyst, PlayStation → https://lnkd.in/efmfNVfM 7. Data Analyst, Live Ops, Mattel → https://lnkd.in/eUQfw3vR 8. Data Analyst, Bristol Myers Squibb → https://lnkd.in/efppbuFk 9. Data Analytics Engineer, Allstate → https://lnkd.in/e5xMZD3s 10. Data Analyst II, Abbott → https://lnkd.in/eBY_CpH2 ♻️ Repost to spread the word PS - Know of any great open roles? Post them in the comments!

Matt MikeOct 5, 2026

A chart can be misleading and accurate at the same time. It all comes down to how it's visualized. Sometimes people do it on purpose, sometimes not, but things like axis and scale can completely change the interpretation of a visual. Here are 5 ways that happens: 𝟭. 𝗧𝗿𝘂𝗻𝗰𝗮𝘁𝗲𝗱 𝗮𝘅𝗶𝘀 → Start the y-axis at 97 instead of 0 and a 5% gap looks like a landslide. 𝟮. 𝗖𝗵𝗲𝗿𝗿𝘆-𝗽𝗶𝗰𝗸𝗲𝗱 𝗿𝗮𝗻𝗴𝗲 → Show 13 months of a 48-month decline and it looks like growth. Pick the window and you can show either direction. 𝟯. 𝗠𝗶𝘀𝘀𝗶𝗻𝗴 𝗱𝗲𝗻𝗼𝗺𝗶𝗻𝗮𝘁𝗼𝗿 → Raw counts put the biggest groups on top. Divide by population and the ranking can flip completely. 𝟰. 𝗗𝘂𝗮𝗹 𝗮𝘅𝗶𝘀 → Give two series their own scales and anything can look related. On one shared scale, ad spend triples while sales barely move. 𝟱. 𝗔𝘃𝗲𝗿𝗮𝗴𝗲 𝗼𝗻𝗹𝘆 → One average can describe two groups that share nothing. Plot every point and the average lands in the gap where nobody is. Both charts in every pair use the exact same numbers. Only the drawing changed. That makes all the difference in how it's interpreted. Never take a visual at face value. Always take a closer look. PS - Which of these do you see most often? _____ ♻️ Repost for the homies 💾 Save for later

Matt MikeOct 7, 2026

AI can build your Power BI reports now. Up until recently, the limitation was creating visuals. Not anymore. Microsoft released Report Authoring skills that let AI coding agents like Claude Code and GitHub Copilot plan, build and publish Power BI reports using natural language. It works in two parts: → 𝗦𝗸𝗶𝗹𝗹𝘀 are instruction files packed with Power BI best practices. They tell the agent what to do. → 𝗧𝗼𝗼𝗹𝘀 are the MCP server and CLIs that read your model and edit the report files. One plugin installs both. Here's how to set it up and start using it. 𝟭. 𝗚𝗲𝘁 𝗿𝗲𝗮𝗱𝘆 → You need an AI coding tool, Node.js 18 or later, and Power BI Desktop to preview the result. Copilot CLI also needs PowerShell 7. 𝟮. 𝗜𝗻𝘀𝘁𝗮𝗹𝗹 𝘁𝗵𝗲 𝗽𝗹𝘂𝗴𝗶𝗻 → Run these two commands in your agent's chat: - /plugin marketplace add microsoft/skills-for-fabric - /plugin install powerbi-authoring@fabric-collection - Then restart and type /skills to confirm they're loaded. 𝟯. 𝗦𝗮𝘃𝗲 𝗮𝘀 𝗮 𝗣𝗕𝗜𝗣 𝗽𝗿𝗼𝗷𝗲𝗰𝘁 → The skills only edit PBIP files in PBIR format. Commit a Git baseline before the agent starts so you can roll back anything you don't like. 𝟰. 𝗔𝘀𝗸 𝗶𝗻 𝗻𝗮𝘁𝘂𝗿𝗮𝗹 𝗹𝗮𝗻𝗴𝘂𝗮𝗴𝗲 → Name the pages, visuals and measures you want. Something like "Add KPI cards and a date slicer to the Overview page." From there, every report moves through 5 stages: plan, design, author, validate and manage. A new report starts at plan. A quick edit jumps straight to author. Once it's running, it can also: → 𝗥𝗲-𝘁𝗵𝗲𝗺𝗲 a whole report in one ask → 𝗠𝗼𝗱𝗲𝗿𝗻𝗶𝘇𝗲 legacy cards and matrices → 𝗗𝗲𝗯𝘂𝗴 visuals stuck on an error icon → 𝗣𝘂𝗯𝗹𝗶𝘀𝗵 straight to a Power BI workspace Overall, it's pretty awesome. I've been playing with these a ton over the last few weeks. PS - Have you tried these new skills yet? _____ ♻️ Repost for the homies 💾 Save for later