Home > POC Agenda > Demo Runbook

🎰 Casino/Gaming Microsoft Fabric POC - Demo Runbook¶

Last Updated: 2026-04-15 | Version: 2.0 Status: ✅ Final | Maintainer: Documentation Team

Presenter Guide for Live Demonstrations

Duration: 90 minutes

📋 Table of Contents¶

• 1. Pre-Demo Checklist
• 2. Demo Flow Overview (90-minute version)
• 3. Demo Sections
• A. Opening (5 min)
• B. Medallion Architecture Demo (25 min)
• C. Real-Time Intelligence Demo (20 min)
• D. Power BI & Direct Lake Demo (20 min)
• E. Governance with Purview Demo (15 min)
• F. Wrap-up & Q&A (5 min)
• 4. Troubleshooting Quick Reference
• 5. Backup Demo Paths
• 6. Appendix

✅ 1. Pre-Demo Checklist¶

1.1 Environment Verification (30 minutes before demo)¶

1.2 Data Verification¶

Quick Validation Queries:

-- Run in lh_bronze SQL endpoint
SELECT 'bronze_slot_telemetry' as table_name, COUNT(*) as row_count FROM bronze_slot_telemetry
UNION ALL
SELECT 'bronze_player_profile', COUNT(*) FROM bronze_player_profile
UNION ALL
SELECT 'bronze_financial_txn', COUNT(*) FROM bronze_financial_txn;

1.3 Browser Tabs to Have Open¶

Open these tabs in order (left to right):

1.4 Pre-Demo Setup Actions¶

1.5 Backup Plans for Common Issues¶

🎬 2. Demo Flow Overview (90-minute version)¶

+-----+------------------+------------------+---------------------+------------------+-----------+-----+
| 0   | 5                | 30               | 50                  | 70               | 85        | 90  |
+-----+------------------+------------------+---------------------+------------------+-----------+-----+
|  A  |       B          |        C         |         D           |        E         |     F     |
|     |                  |                  |                     |                  |           |
|OPEN | MEDALLION ARCH   | REAL-TIME        | POWER BI &          | PURVIEW          | WRAP-UP   |
|     | Bronze/Silver/   | Eventhouse       | DIRECT LAKE         | GOVERNANCE       | Q&A       |
|     | Gold Demo        | Streaming Demo   | Dashboards          | Lineage          |           |
+-----+------------------+------------------+---------------------+------------------+-----------+-----+
  5m         25m                20m                  20m                  15m             5m

Timing Summary¶

🎯 3. Demo Sections¶

A. Opening (5 min)¶

Objective: Set context, establish credibility, outline what attendees will see.

A.1 Welcome and Context Setting (2 min)¶

[Show: Architecture slide or README in GitHub]

Talking Points:

"Welcome to the Microsoft Fabric Casino & Gaming POC demonstration. Today, we'll walk through a complete end-to-end analytics platform purpose-built for the gaming industry.

This isn't just a toy demo - what you'll see represents patterns you can deploy in production to address real challenges: - Real-time floor monitoring - Regulatory compliance (NIGC MICS, FinCEN) - Player analytics and churn prevention - Sub-second business intelligence"

A.2 Architecture Overview (2 min)¶

[Show: Architecture diagram from README or ARCHITECTURE.md]

Navigation: 1. Open browser Tab 1 (Fabric Portal) 2. Navigate to: Workspace > Show workspace overview

Talking Points:

"Let me show you the architecture we've built:

Data flows from left to right: - Casino floor systems (slot machines, table games, player cards) - Through our Medallion architecture (Bronze for raw, Silver for cleansed, Gold for business-ready) - Into two consumption paths: batch analytics via Direct Lake Power BI, and real-time via Eventhouse - All governed through Microsoft Purview

The beauty of Fabric is that all of this lives in ONE platform - no data movement between services."

[Screenshot: Architecture Overview]

┌─────────────────────────────────────────────────────────────────┐
│                                                                 │
│   [Place screenshot of workspace overview here]                 │
│                                                                 │
└─────────────────────────────────────────────────────────────────┘

A.3 POC Objectives (1 min)¶

Talking Points:

"By the end of this demonstration, you'll see: 1. How the Medallion architecture transforms raw casino data into business insights 2. Real-time slot floor monitoring with sub-second latency 3. Executive dashboards powered by Direct Lake - no data refresh needed 4. Complete data lineage and governance through Purview

Let's dive in."

Transition:

"We'll start where all good data platforms start - with the data. Let me show you the Medallion architecture in action."

B. Medallion Architecture Demo (25 min)¶

Objective: Demonstrate the Bronze/Silver/Gold pattern with real casino data.

B.1 Bronze Layer - Raw Data Landing (8 min)¶

B.1.1 Navigate to Bronze Lakehouse¶

Navigation: 1. Browser Tab 2 (or navigate in Tab 1) 2. Click: Workspace > lh_bronze Lakehouse 3. Expand: Tables folder

[Screenshot: Lakehouse Explorer - Bronze Tables]

Source: Microsoft Fabric Lakehouse Overview

Talking Points:

"This is our Bronze layer - the landing zone for all raw data. Notice we have tables for each data domain: - Slot telemetry from the gaming floor - Player profiles from the loyalty system - Financial transactions from the cage - Compliance filings for regulatory reporting"

B.1.2 Explore Raw Data¶

Navigation: 1. Click: bronze_slot_telemetry table 2. Click: "Preview" button

Talking Points:

"Let's look at the raw slot data. This comes directly from the SAS protocol on the gaming machines.

Key things to notice: - We preserve the data exactly as received - no transformation yet - Every record has metadata: ingestion timestamp, batch ID, source file - This allows us to reprocess if business rules change

In a typical casino, we're ingesting millions of these events per day."

Navigation: 1. Click: bronze_player_profile table 2. Click: "Preview"

Talking Points:

"Here's our player data. Notice something important - we DON'T see raw SSN values here. Even at the Bronze layer, we hash PII on ingestion.

This is a compliance requirement - we never store unprotected social security numbers. The original value is hashed using SHA-256 before landing in the lakehouse."

B.1.3 Show Data Volume¶

Navigation: 1. Right-click bronze_slot_telemetry 2. Select: "View table details" or run quick SQL

SQL to run (if needed):

SELECT
    COUNT(*) as total_records,
    MIN(event_timestamp) as earliest_event,
    MAX(event_timestamp) as latest_event,
    COUNT(DISTINCT machine_id) as unique_machines
FROM bronze_slot_telemetry;

Talking Points:

"You can see we have [X] records spanning [Y] days of gaming activity across [Z] machines. This is realistic scale for a medium-sized casino floor."

Transition:

"Raw data isn't useful for analytics. Let's see how we transform it in the Silver layer."

B.2 Silver Layer - Data Quality & Cleansing (8 min)¶

B.2.1 Navigate to Silver Lakehouse¶

Navigation: 1. Click: Workspace breadcrumb 2. Click: lh_silver Lakehouse 3. Expand: Tables folder

[Screenshot: Lakehouse Explorer - Silver Tables]

Talking Points:

"The Silver layer is where we apply business logic. Notice the tables have different names - they're cleansed, validated versions of the raw data."

B.2.2 Execute Silver Transformation Notebook¶

Navigation: 1. Click: Workspace breadcrumb 2. Navigate to: Notebooks folder 3. Open: 01_silver_slot_cleansed

Talking Points:

"Let me show you a transformation notebook. This is PySpark running natively in Fabric."

Walk through the notebook (do NOT run live - show code):

Cell 1 - Configuration:

"First, we read from Bronze. No credentials needed - Fabric handles authentication automatically."

Cell 2 - Data Quality Rules:

"Here's where we apply business rules: - Remove records where machine_id is null - these are orphan events - Fix negative coin values - data quality issues from the source - Deduplicate based on machine, timestamp, and event type"

Cell 3 - Quality Score:

"We calculate a data quality score for each record. This lets us track data health over time and identify problematic sources."

[Screenshot: Notebook code showing quality rules]

B.2.3 Show Data Quality Improvements¶

Navigation: 1. Return to lh_silver Lakehouse 2. Open SQL endpoint 3. Run quality comparison query

SQL Query:

SELECT
    'Bronze' as layer,
    COUNT(*) as total_records,
    COUNT(CASE WHEN machine_id IS NULL THEN 1 END) as null_machines,
    COUNT(CASE WHEN coin_in < 0 THEN 1 END) as negative_values
FROM lh_bronze.bronze_slot_telemetry
UNION ALL
SELECT
    'Silver',
    COUNT(*),
    COUNT(CASE WHEN machine_id IS NULL THEN 1 END),
    COUNT(CASE WHEN coin_in < 0 THEN 1 END)
FROM lh_silver.silver_slot_cleansed;

Talking Points:

"Look at the difference. Bronze has [X] null machines and [Y] negative values. Silver has zero of both. The data is now validated and ready for analytics."

Transition:

"Clean data is great, but business users need metrics and KPIs. That's what the Gold layer provides."

B.3 Gold Layer - Business-Ready Analytics (9 min)¶

B.3.1 Navigate to Gold Lakehouse¶

Navigation: 1. Click: Workspace breadcrumb 2. Click: lh_gold Lakehouse 3. Expand: Tables folder

[Screenshot: Lakehouse Explorer - Gold Tables]

Talking Points:

"The Gold layer is optimized for business consumption. Notice these are aggregated tables: - gold_slot_performance - machine-level KPIs by day - gold_player_360 - complete player profiles with lifetime value - gold_compliance_reporting - regulatory filing summaries"

B.3.2 Execute Gold Aggregation Notebook¶

Navigation: 1. Navigate to Notebooks 2. Open: 01_gold_slot_performance

Walk through key sections:

KPI Calculations:

"Here we calculate casino-specific metrics: - Net Win = Coin In minus Coin Out - this is gross gaming revenue - Hold Percentage = Net Win divided by Coin In - the actual house edge - Theoretical Win = Expected win based on game math (typically 8%) - Hold Variance = Actual versus expected - this alerts us to problems"

Talking Points:

"If a machine's hold variance is significantly negative, it could indicate: - A malfunction paying out too much - Advantage play or fraud - Math model errors

These are the metrics that floor managers and regulators care about."

B.3.3 Show Business-Ready Data¶

Navigation: 1. Return to lh_gold Lakehouse 2. Click: gold_slot_performance 3. Preview the data

Talking Points:

"This is what your BI team consumes. Clean, aggregated, with all KPIs pre-calculated.

Notice it's partitioned by business_date - this optimizes query performance when slicing by time period, which is 90% of gaming analytics."

SQL Query (optional):

SELECT
    business_date,
    SUM(total_coin_in) as total_coin_in,
    SUM(net_win) as total_net_win,
    AVG(actual_hold_pct) as avg_hold_pct
FROM gold_slot_performance
GROUP BY business_date
ORDER BY business_date DESC
LIMIT 7;

"Here's a week of gaming revenue at a glance. This query runs in under a second against millions of records."

Transition:

"So far, we've seen batch processing. But casinos need real-time visibility into their floor. Let me show you Real-Time Intelligence."

C. Real-Time Intelligence Demo (20 min)¶

Objective: Demonstrate live streaming analytics with Eventhouse and KQL.

C.1 Eventhouse Setup Overview (5 min)¶

C.1.1 Navigate to Eventhouse¶

Navigation: 1. Click: Workspace breadcrumb 2. Click: eh_casino_realtime Eventhouse 3. Click: casino_floor_monitoring database

[Screenshot: Eventhouse with KQL Database]

Talking Points:

"Real-Time Intelligence in Fabric is built on the same engine as Azure Data Explorer - incredibly fast for time-series data.

This Eventhouse receives streaming events from the casino floor. Data arrives within seconds of a slot pull or jackpot hit."

C.1.2 Show Eventstream Configuration¶

Navigation: 1. Click: Workspace breadcrumb 2. Navigate to: es_slot_telemetry Eventstream 3. Show the data flow diagram

[Screenshot: Eventstream visual designer]

Source: Create an Eventstream in Microsoft Fabric

Talking Points:

"Here's how data flows in: - Source: Event Hub receiving slot machine events (could also be IoT Hub, Kafka, custom app) - Transformations: Parse JSON, add timestamp - Destination: Our KQL database

This is no-code ingestion - configured entirely through the visual designer."

C.2 Start Streaming Producer (if not already running)¶

Navigation: 1. Open terminal/command prompt (can be hidden if pre-started) 2. Verify producer is running

Talking Points:

"In a real casino, the slot management system pushes events continuously. For this demo, we have a simulator generating realistic telemetry."

Note: If starting the producer live:

cd data_generation
python generators/streaming/event_hub_producer.py

C.3 Live Data in KQL (10 min)¶

C.3.1 Navigate to KQL Query Editor¶

Navigation: 1. Return to Eventhouse 2. Click: "Explore your data" or open KQL Queryset

[Screenshot: KQL Query Editor]

C.3.2 Real-Time Floor Summary¶

Run KQL Query:

// Current floor activity (last 5 minutes)
SlotEvents
| where event_timestamp > ago(5m)
| summarize
    active_machines = dcount(machine_id),
    total_coin_in = sum(coin_in),
    total_games = sum(games_played),
    jackpots = countif(event_type == "JACKPOT")
by zone
| order by total_coin_in desc

Talking Points:

"This shows real-time activity across the casino floor. In the last 5 minutes: - [X] active machines across [Y] zones - $[Z] in coin-in (wagers) - [N] jackpots hit

A floor manager can see at a glance which zones are hot."

C.3.3 Machine Anomaly Detection¶

Run KQL Query:

// Machines with unusual hold (last hour)
SlotEvents
| where event_timestamp > ago(1h)
| summarize
    coin_in = sum(coin_in),
    coin_out = sum(coin_out)
by machine_id, zone
| extend
    net_win = coin_in - coin_out,
    hold_pct = (coin_in - coin_out) / coin_in * 100
| where hold_pct < 2 or hold_pct > 15
| project machine_id, zone, coin_in, net_win, hold_pct
| order by hold_pct asc

Talking Points:

"Here's where it gets interesting for operations. We're looking for machines outside normal parameters: - Hold below 2% - machine might be malfunctioning or being exploited - Hold above 15% - could indicate a stuck mechanism or player disputes

These alerts would typically trigger a floor check or technician dispatch."

C.3.4 Jackpot Tracking¶

Run KQL Query:

// Recent jackpots
SlotEvents
| where event_type == "JACKPOT"
| where event_timestamp > ago(1h)
| project
    event_timestamp,
    machine_id,
    zone,
    jackpot_amount,
    player_id
| order by event_timestamp desc
| take 10

Talking Points:

"Every jackpot is tracked in real-time. For hand-pays over $1,200, we need to generate W-2G tax forms. This data feeds directly into our compliance reporting."

C.4 Real-Time Dashboard Updates (5 min)¶

C.4.1 Navigate to Real-Time Dashboard¶

Navigation: 1. Click: Workspace breadcrumb 2. Navigate to: Casino Floor Monitor dashboard

[Screenshot: Real-Time Dashboard with multiple tiles]

Talking Points:

"This dashboard auto-refreshes every 30 seconds. Let's watch the numbers change..."

Wait for refresh to show updated values

"There - you can see the coin-in numbers just updated. This is the same view your floor managers would have on monitors throughout the casino."

C.4.2 Highlight Key Tiles¶

Walk through dashboard tiles:

Talking Points:

"Notice the alert panel - it's automatically flagging: - Hot machines (unusually high hold) - Big jackpots requiring floor attention - Silent machines that stopped reporting

This is proactive operations - problems are surfaced before customers complain."

Transition:

"Real-time is critical for operations, but executives and analysts need historical trends and deep analysis. Let me show you our Power BI integration."

D. Power BI & Direct Lake Demo (20 min)¶

Objective: Demonstrate sub-second analytics with Direct Lake and interactive dashboards.

D.1 Open Semantic Model (5 min)¶

D.1.1 Navigate to Semantic Model¶

Navigation: 1. Click: Workspace breadcrumb 2. Click: Casino Analytics Model semantic model 3. Show model view

[Screenshot: Semantic Model in diagram view]

Talking Points:

"This is our Direct Lake semantic model. A few important things:

1. Storage mode is Direct Lake - data is read directly from Delta tables in OneLake
2. No scheduled refresh needed - when Gold layer updates, reports see it immediately
3. Sub-second queries - the Fabric engine optimizes for BI workloads"

D.1.2 Show Direct Lake Connection¶

Navigation: 1. Click: Model settings (gear icon) 2. Show storage mode setting

Source: Direct Lake Overview

Talking Points:

"See here - Storage mode: Direct Lake. This is different from: - Import mode (copies data into Power BI) - DirectQuery (queries source on every interaction)

Direct Lake gives you import-mode performance with DirectQuery-style freshness. Best of both worlds."

D.2 Navigate Executive Dashboard (8 min)¶

D.2.1 Open Power BI Report¶

Navigation: 1. Browser Tab 4 (Power BI) 2. Or navigate: Workspace > Executive Dashboard report

[Screenshot: Executive Dashboard full view]

Talking Points:

"Here's what the casino GM sees every morning. One screen, all key metrics."

D.2.2 Walk Through KPI Cards¶

Point to each card:

Talking Points:

"These are live numbers from Gold layer. If a notebook runs and updates Gold tables, these cards reflect it within minutes."

D.2.3 Show Net Win Trend¶

Click on the trend line chart

Talking Points:

"This trend shows daily net win with comparison to prior period. Notice: - Weekends peak (Friday-Saturday) - Slight dip mid-week - This is typical casino seasonality

The dotted line shows prior month for context."

D.2.4 Zone Performance Bar Chart¶

Hover over zones

Talking Points:

"Here's performance by zone. VIP zone generates disproportionate revenue - that's expected given higher limits.

But notice Penny Palace - high volume, lower revenue per transaction. Different player profile."

D.3 Drill-Through Examples (4 min)¶

D.3.1 Drill from Zone to Machines¶

Navigation: 1. Right-click on a zone bar 2. Select: Drill through > Machine Details

Talking Points:

"Let's drill into the VIP zone... Here are the individual machines, ranked by net win.

This helps identify: - Top performers to protect - Underperformers that might need different placement or removal"

D.3.2 Show Cross-Filtering¶

Navigation: 1. Click on a date in the trend chart 2. Show how other visuals filter

Talking Points:

"Power BI cross-filtering in action. Click one date, everything else filters. Want to see what happened on New Year's Eve? One click."

D.4 Q&A Visual Demo (3 min)¶

D.4.1 Navigate to Q&A Page (if available)¶

Navigation: 1. Navigate to Q&A visual or page 2. Or add Q&A visual to demo

Type natural language query:

"Show me net win by zone for last week"

Talking Points:

"Power BI natural language query - ask business questions in plain English.

'Show me net win by zone for last week'

The model understands casino terminology because we defined glossary terms and proper measure names."

Try another query:

"Which machines have the highest hold percentage?"

Talking Points:

"This is how you democratize analytics. Business users don't need to know DAX or SQL - they just ask questions."

Transition:

"Beautiful dashboards are great, but in regulated industries, you need to prove data integrity and lineage. Let me show you Purview."

E. Governance with Purview Demo (15 min)¶

Objective: Demonstrate data governance, lineage, and compliance capabilities.

E.1 Data Catalog Search (4 min)¶

E.1.1 Navigate to Purview Portal¶

Navigation: 1. Browser Tab 5 (Purview Portal) 2. Or navigate to: purview.microsoft.com

[Screenshot: Purview Portal home]

Source: Microsoft Purview Governance Portal

Talking Points:

"Microsoft Purview is our governance hub. Think of it as the 'Google for your data' - catalog everything, track lineage, enforce policies."

E.1.2 Search for Casino Data¶

Navigation: 1. Click: Data Catalog (or search bar) 2. Search: "slot"

[Screenshot: Search results showing slot-related assets]

Talking Points:

"Search for 'slot' and we find all assets related to slot machines: - Bronze raw tables - Silver cleansed tables - Gold aggregations - Even the Power BI reports

Everything is automatically cataloged when we scan Fabric."

E.1.3 Show Asset Details¶

Navigation: 1. Click on gold_slot_performance in results 2. Show asset detail page

Talking Points:

"Each asset has: - Schema with column descriptions - Classifications (auto-detected PII, financial data) - Ownership and stewardship - Related glossary terms"

E.2 Lineage Visualization (5 min)¶

E.2.1 Navigate to Lineage View¶

Navigation: 1. On asset page, click: "Lineage" tab 2. Show end-to-end data flow

[Screenshot: Lineage diagram showing Bronze -> Silver -> Gold -> Power BI]

Talking Points:

"This is the killer feature for compliance. Complete lineage from source to report.

Follow the path: 1. bronze_slot_telemetry - raw data lands here 2. Notebook transformation - cleansing logic 3. silver_slot_cleansed - validated data 4. Another notebook - aggregation 5. gold_slot_performance - business metrics 6. Casino Analytics Model - semantic layer 7. Executive Dashboard - final report

When a regulator asks 'where did this number come from?' - one click shows the complete journey."

E.2.2 Impact Analysis¶

Navigation: 1. Click on a source table 2. Show downstream dependencies

Talking Points:

"Now the power move - impact analysis. Click on Bronze table and see: - What Silver tables depend on it? - What Gold tables? - Which reports?

If you need to change a source schema, you know exactly what breaks downstream."

E.3 Sensitivity Labels (3 min)¶

E.3.1 Show Classifications¶

Navigation: 1. Navigate to silver_player_master asset 2. Show column classifications

[Screenshot: Asset showing PII classifications]

Talking Points:

"Purview automatically classifies sensitive data: - email - classified as Email Address - phone - classified as Phone Number - date_of_birth - Date of Birth - ssn_hash - Government ID (hashed)

These classifications drive downstream policies - who can see what."

E.3.2 Data Protection Policies¶

Talking Points:

"With sensitivity labels, you can enforce: - Row-level security based on user role - Column masking for PII - Export restrictions for highly confidential data

This is how you prove to regulators that you're protecting player data."

E.4 Glossary Terms (3 min)¶

E.4.1 Navigate to Glossary¶

Navigation: 1. Click: Data catalog > Glossary 2. Navigate to: Casino Operations hierarchy

[Screenshot: Business glossary with casino terms]

Talking Points:

"The business glossary ensures everyone speaks the same language. Look at our casino metrics:

Coin In - Total amount wagered by players Hold Percentage - Actual house edge Theoretical Win - Expected win based on game math

These definitions link to actual data columns. When a BI developer sees 'hold_pct', they click through to see the official definition."

E.4.2 Show Term Linked to Asset¶

Navigation: 1. Click on "Net Win" glossary term 2. Show linked assets

Talking Points:

"The glossary term 'Net Win' is linked to: - The gold table column - The DAX measure - Related documentation

No more confusion about what metrics mean."

Transition:

"That concludes the main demo. Let me summarize what we've seen and open it up for questions."

F. Wrap-up & Q&A (5 min)¶

Objective: Reinforce key messages, provide next steps, facilitate questions.

F.1 Summary of Capabilities (2 min)¶

Return to workspace overview or architecture slide

Talking Points:

"In 85 minutes, we've seen a complete modern data platform:

1. Medallion Architecture - Bronze for raw data with full auditability - Silver for data quality and validation - Gold for business-ready analytics

2. Real-Time Intelligence - Sub-second streaming from the casino floor - KQL for operational monitoring - Auto-refreshing dashboards

3. Direct Lake Power BI - No-refresh business intelligence - Interactive drill-through - Natural language Q&A

4. Purview Governance - Complete data lineage - Automatic PII classification - Business glossary

All in ONE platform. No data movement. Unified security."

F.2 Next Steps Discussion (2 min)¶

Talking Points:

"Where do we go from here?

Immediate next steps: - Connect your actual data sources - Customize the notebooks for your business rules - Train your team on the tutorials we provide

This repository includes: - 9 step-by-step tutorials - Sample data generators - Infrastructure-as-Code for deployment - Full documentation

The code is yours - fork it, customize it, run with it."

F.3 Resource Links (30 sec)¶

Show or verbally provide:

F.4 Q&A (remaining time)¶

Talking Points:

"I'm happy to take questions. What would you like to dive deeper into?"

Common questions to anticipate:

🔧 4. Troubleshooting Quick Reference¶

🔀 5. Backup Demo Paths¶

5.1 If Bronze Lakehouse Not Available¶

Alternative: Show Architecture + Screenshots

1. Display architecture diagram from docs/ARCHITECTURE.md
2. Walk through Bronze layer concepts verbally
3. Show sample data files in sample-data/ folder

Talking Points:

"While we work on the environment issue, let me show you what the Bronze layer contains conceptually..."

5.2 If Real-Time Streaming Not Working¶

Alternative: Historical KQL Demo

1. Navigate to Eventhouse
2. Run queries against historical data (remove time filters)

Modified Query:

// Use historical data instead of real-time
SlotEvents
| take 1000
| summarize
    active_machines = dcount(machine_id),
    total_coin_in = sum(coin_in)
by zone

Talking Points:

"Normally this would show live data. Let me show you the same patterns against our historical dataset..."

5.3 If Power BI Not Loading¶

Alternative: PDF Export + DAX Discussion

1. Open pre-exported PDF of dashboards
2. Walk through DAX measures in semantic model

Pre-export these reports: - Executive Dashboard (full) - Slot Operations Report - Player Analytics Report

Talking Points:

"While Power BI initializes, let me walk you through the report design using this export..."

5.4 If Purview Not Accessible¶

Alternative: Use Mermaid Lineage Diagram

1. Open docs/ARCHITECTURE.md
2. Show lineage section with Mermaid diagram

Talking Points:

"Purview provides this visually in an interactive way. Here's the same lineage concept in a diagram..."

bronze_slot_telemetry --> silver_slot_cleansed --> gold_slot_performance --> Power BI

5.5 Shortened Demo (45 minutes)¶

If time is cut, prioritize:

Skip: - Bronze/Silver detailed walkthrough - Notebook execution - KQL query writing - Q&A visual demo

📎 6. Appendix¶

6.1 Key URLs¶

6.2 Login Credentials Reference¶

SECURITY NOTE: Never include actual credentials in documentation. Use placeholders and a secure credential manager.

Demo Account (if applicable):

Username: [PLACEHOLDER - ENTER BEFORE DEMO]
Workspace: casino-fabric-poc
Capacity: F64

6.3 Emergency Contacts¶

6.4 Quick KQL Reference (for ad-hoc queries)¶

// Last N events
SlotEvents | take 100

// Filter by time
SlotEvents | where event_timestamp > ago(5m)

// Aggregate
SlotEvents | summarize count() by zone

// Join
SlotEvents
| join kind=inner (PlayerProfiles) on player_id

// Render visualization
SlotEvents | summarize count() by bin(event_timestamp, 1h) | render timechart

6.5 Quick DAX Reference (for ad-hoc measures)¶

// Basic sum
Total Coin In = SUM(gold_slot_performance[total_coin_in])

// Division with error handling
Hold % = DIVIDE([Net Win], [Total Coin In], 0) * 100

// Time intelligence
Coin In MTD = TOTALMTD([Total Coin In], dim_date[date_key])

// Conditional count
VIP Players = CALCULATE(COUNTROWS(gold_player_360), gold_player_360[tier] = "VIP")

6.6 Demo Feedback Form Link¶

After the demo, share this link for feedback:

[INSERT FEEDBACK FORM URL]

📜 Version History¶

📖 Related Documents¶