Let me start by loading the relevant skills and understanding the data.
Now let me read the slide reference and semantic-sql skill before proceeding.
Great, I have a thorough understanding of the data and the workflow. Let me start the analysis. This is a customer survey dataset — I'll explore it systematically.
Phase 1 — Explore the data
Excellent! Now I have a comprehensive picture of the data. Let me proceed to Phase 2 — Generate charts for the PPT report.
I need to use member short names. Let me check the error and fix the aliases. The issue is that SQL aliases must match Cube member short names from the schema. Let me fix
All 10 charts generated successfully. Now let me write the summary and outline, then proceed with the PPT pipeline.
Now I need to ask the user to choose a template for the PPT.
The user selected the dark-green template. Let me proceed with the report pipeline.
Guideline succeeded. Now I proceed directly to generating the presentation
Let me present the final deliverable:
Bayes: Here's your Customer Survey Analysis Report — a 13-slide presentation covering the full analysis.
Report Summary
1,000 respondents surveyed globally across 5 regions, with ratings on satisfaction, NPS, support, usability, and more.
Key Findings in the Report
| Finding | Detail |
|---|---|
| Overall Satisfaction | 7.51/10 — moderate, with room to grow |
| NPS Split | 33.4% Promoters, 38.3% Passives, 28.3% Detractors |
| Weakest Areas | Customer Support (1.69/10) & Ease of Use (1.76/10) |
| Mobile Dominance | 59.7% of users are on mobile phones |
| Issues | 67.6% reported no issues; usability (11.2%) and technical (10.5%) are top complaints |
| Renewal Risk | Less than 2% expressed likely renewal intent across all regions |
| Regional Leader | North America (7.64 satisfaction); Asia Pacific trails (7.44) |
The presentation is saved as Customer_Survey_Analysis.pres — you can download and view it directly.
Let me load the dashboard skill to build an interactive BI dashboard.
Now let me load the design template and reference files.
Now let me explore the anchor cube to get the exact field IDs I need for the charts and dashboard
Now let me read the full explore output to get exact field IDs
Now let me create the dashboard design doc and then the chart JSON files. I need to create new charts for the dashboard — I'll create KPI big number charts and a few others. Let me need to read the embed-chart skill for the chart JSON format
Now let me build the dashboard. First, the design doc, then charts, then dashboard JSON. I'll create the design doc and get the chart samples.
Now I have all the reference materials. Let me write the dashboard design doc first, then create all charts, then the dashboard JSON.
First, let me create all the charts I need. I'll reuse the save_embed_chart charts from the analysis that already exist, and create new ones for KPIs we need for the dashboard.
Let me check what charts already exist from the analysis step
Now I need to create the 4 KPI big number charts that are new (not from the analysis). Let me create them using save_embed_chart since they're simple queries
Now let me check what chart files we have on disk from the prior analysis + the new ones
All charts are available. Now let me build the dashboard JSON. I need to check the exact chart UUIDs from the chart JSON files first.
Now let me build the dashboard JSON with the correct UUIDs, proper field IDs, and filters.
Dashboard verified successfully! Let me report it.
Bayes: Your Customer Survey Analysis Dashboard is ready!
Dashboard Structure
The dashboard is a single-page interactive BI dashboard with 15 tiles organized in 3 sections:
1. Overview (Hero KPIs + NPS)
| Tile | Content |
|---|---|
| 4 KPI cards | Avg Satisfaction (7.51), Avg NPS (7.53), Total Respondents (1,000), Avg Age (39.2) |
| NPS Pie | Promoter / Passive / Detractor split |
| Region Bar | Satisfaction & NPS across 5 regions |
2. Key Insights
Actionable bullets highlighting the weakest dimensions (Support 1.69, Ease of Use 1.76), mobile dominance (60%), and the NPS risk profile.
3. Demographics & Usage
| Tile | Content |
|---|---|
| Gender comparison | Satisfaction, NPS, and recommend intent by gender |
| Device Type | Mobile vs Desktop vs Tablet pie |
| Primary Use Case | Entertainment, Personal, Education, Work, All |
| Issues Experienced | Usability, Technical, Billing issues breakdown |
Filters
- Region filter — applies to all charts
- Gender filter — applies to all charts
The dashboard file is saved as dashboards/customer_survey_dashboard.json — ready to embed and explore interactively.
