Video summary

[GLOBAL] - Career Boost with Power BI and AI - Day #04

Main summary

Key takeaways

Technology

Technological focus / key concepts (Day #04)

  • Data analysis with AI applied to dashboards and reporting as the “next step” after:
    • Day 1: productivity with AI (prompting and using tools like ChatGPT / Cloud / Copilot / Gemini in similar workflows)
    • Day 2: Power BI data transformation, modeling, visualization, and publishing
    • Day 3: AI-assisted “vibe coding” tools (prototype creation via platforms like lovable / manus)
  • Core message: integrating AI tools with Power BI + Excel changes how professionals interact with data—toward faster, more interactive, and more conversational analytics.

Tools and product features demonstrated

1) Microsoft Copilot in Excel (Copilot 365 / M365 Copilot)

Copilot requires that:

  • the Excel file is autosaved
  • the organization/IT allows Copilot to be used

In a spreadsheet case (sales analytics), Copilot:

  • created a new worksheet/tab automatically (e.g., “sales analysis”)
  • generated tables, calculations, and charts
  • answered a set of “four questions” posed by a manager/boss (sales analytics tasks)

Validation emphasized:

  • results were checked against Power BI comparisons and numbers matched
  • verifying that cells contain real formulas (not hardcoded pasted values)

Enhancement suggestions:

  • visuals and storytelling can be improved further (colors, narrative/executive summary text)

2) Microsoft Copilot “skills”

“Skills” are described as reusable AI instructions—an approach for teaching consistent business/reporting formats.

Example usages mentioned:

  • formatting/standards for recurring analyses (e.g., three-statement financial model)
  • reusable comparable/company analysis patterns
  • reduces the need to rebuild prompts from scratch

3) Cloud.ai integrated inside Excel

Requirements:

  • a Cloud license
  • installing/adding the Excel add-in

Model choice guidance from the instructor:

  • “Fable” may be smart but too expensive for Excel use
  • “Opus” is preferred for strong results at better cost

Cloud was shown performing similar analytics tasks by:

  • modifying/adding helper columns
  • calculating metrics and building charts
  • including/using skills and “reasoning” steps (as depicted during the run)

Comparison goal:

  • side-by-side comparison of Copilot vs Cloud outputs in Excel
  • focus on accuracy, dynamic formulas, and ease/speed for managers

4) Power BI vs AI-in-Excel workflow

The “AI approach” is positioned as:

  • not replacing Power BI entirely, but accelerating the first draft/analysis phase
  • AI can generate analysis artifacts in Excel quickly, then Power BI becomes the final BI/reporting layer

Also demonstrated:

  • Copilot inside Excel can create filters/slicers and narrative elements
  • dynamic updates are important when new monthly data arrives

5) Extra AI capabilities beyond spreadsheets

  • Copilot for open-ended analysis conversations (not only file edits)
    • example: benchmarking (checking external market conditions online)
  • Copilot for presentation outputs
    • producing PowerPoint and shareable images
  • Concept of AI agents
    • forecasting scenario using weather data for a “gelataria” sales plan
    • example of creating a role-specific agent that automates decision support

6) AI agent example via WhatsApp + n8n

An “Exodus Experts PowerBI assistant” concept:

  • WhatsApp messages trigger an AI workflow
  • described as built with n8n nodes/steps
  • includes security/permission checks before answering
  • integrates Power BI service access to respond back with results

Complexity emphasized:

  • requires security/credentials and Power BI service involvement
  • not taught end-to-end in the live session, but demonstrated as a capability

7) Power BI operations: updating dashboards

How to automate monthly updates:

  • publish to Power BI Service
  • configure a gateway to bridge between new monthly files and the dataset
  • schedule refresh frequency (e.g., first day of the month)

Tutorial / workshop workflow structure (what attendees were guided to do)

  • Use the same “day #2” dataset for the repeated business case.

  • Create an Excel-based “sales analysis tool” using AI:

    1. open the Excel file
    2. select Copilot or Cloud
    3. paste “context” + ask AI to produce the report for a manager’s four questions
    4. inspect results:
      • charts match the expected metrics
      • ensure calculations are formulas
      • check dynamic behavior and accuracy

Reviews / guidance / training logistics mentioned

  • Training rewatch access:
    • recorded classes for the week are stated as available for free until Saturday (for the basic course portion)
  • Certification process:
    • a live “certification test” with multiple-choice questions
    • only one attempt
    • certification downloadable after completion
  • Instructor claims about training structure:
    • Power BI is broken down into phases:
      • foundation (extract/transform/load, clean data)
      • pillars (data model: fact/dimension tables, measures/KPIs, visuals, publish)
  • Strong emphasis on employability/career outcomes:
    • improved speed/quality makes professionals look more capable
    • expected impact on visibility, promotions, and job opportunities

Main speakers / sources

  • Speaker: Tarisio (also referenced as “Tarisio” / “Microsoft certified trainer”; instructor and organizer)
  • AI tools featured as sources: Microsoft Copilot (Excel / M365 Copilot) and Cloud.ai (model variants like “Opus”), plus mentions of ChatGPT/Gemini as earlier workflow comparators.

Original video