docs
Data Integration
Using Data in Presentations

Using Data in Presentations

This guide covers how data integrates into your presentation workflow, from the Build phase through to final generation.

Two Ways to Use Data

Octigen supports two approaches to data-driven presentations:

1. Workflow-Based (Pre-Configured)

For templates with data-connected shapes:

  • Data configuration happens before the Build phase
  • All data is validated upfront
  • Charts and tables are populated automatically during generation
  • Best for recurring reports with consistent data sources

2. Ad-Hoc (During Build Phase)

For adding data flexibly during content creation:

  • Add data objects from collections while building content
  • Use the Content Assistant to reference data
  • Incorporate data tables into your content
  • Best for exploratory or one-off presentations

Both approaches can be combined in the same presentation.

Workflow-Based Data Integration

Starting a Data-Connected Presentation

  1. Select your template - Must have data-connected shapes
  2. Choose a workflow - Or select "Create Freely"
  3. Configure data - The Data Configuration Modal appears
  4. Set parameters - Fill in global parameters (dates, IDs, etc.)
  5. Validate - Ensure all data exists
  6. Proceed - Continue to Build phase with validated data

During the Build Phase

After data configuration, you'll see in your content list:

📄 Content Pieces
├── Main Content (editable)
├── [Your other content...]
│
└── 📊 Connected Data (read-only)
    ├── monthly-sales: 2024-12-01 (245 rows × 8 cols)
    └── portfolio-holdings: GROWTH_FUND, 2024-12-01 (128 rows × 12 cols)

Connected data is read-only because:

  • It's sourced from your data collections
  • Parameters are locked to the configured values
  • Changes should be made to source files, not in the presentation

During Structure Phase

The Structure Analysis AI:

  • Knows which shapes are data-connected
  • Plans slide layouts to accommodate charts and tables
  • Ensures data shapes are placed appropriately

During Populate Phase

When slides are populated:

  • Data-connected charts receive their data automatically
  • Tables are filled with the validated data
  • No AI interpretation - direct data transfer
  • Column mappings apply if configured

What Gets Populated

Charts:

  • Data series populated from specified columns
  • Labels from first column (typically)
  • Values from data columns
  • Chart type follows template shape type

Tables:

  • Headers from data column names
  • Rows from data rows
  • Formatting follows template table style

Ad-Hoc Data Integration

Adding Data During Build Phase

Even without pre-configured data connections, you can add data to your content:

  1. Open the add content menu in the Build phase
  2. Select "Add Data Object"
  3. The Select Data Modal opens

The Select Data Modal

This modal shows all available data collections:

📊 Select Data Object

Global Parameters
 * Valuation Date: [📅 2024-12-01        ]
 * Portfolio ID:   [GROWTH_FUND__________]
  ┌─ Monthly Sales ───────────────────────────┐
    Parameters: valuation_date             
    Status: ✅ Resolved (245 rows × 8 cols).
    [Preview Data ▼]  [Add to Content]
  └───────────────────────────────────────────┘

  ┌─ Portfolio Holdings ──────────────────────┐
    Parameters: portfolio_id, valuation_date
    Status: ⏳ Not resolved
    [Resolve Data]
  └───────────────────────────────────────────┘

Resolution Flow

  1. Enter global parameters at the top
  2. Expand a collection to see its specific parameters
  3. Click "Resolve Data" to fetch the data
  4. Preview the data to verify it's correct
  5. Click "Add to Content" to add it as a content piece

Data as Content Pieces

When you add data via the ad-hoc modal:

  • A new content piece is created
  • Contains the data in table format
  • Named with collection and parameters (e.g., "monthly-sales: 2024-12-01")
  • Available for AI to reference during structuring and population

Using with Content Assistant

You can reference added data in prompts:

"Create a summary of the sales trends based on the monthly-sales data"

"Build a narrative around the top 10 holdings from the portfolio data"

"Compare the December and November figures from the market data"

The Content Assistant can read and interpret your data to create content.

Data in Different Presentation Modes

Freeform Mode

In freeform (no workflow) mode:

  • No pre-configuration of data connections
  • Add data ad-hoc during Build phase
  • AI uses data as reference for content creation
  • Data tables can inform chart suggestions

Workflow Mode

In workflow mode with data-connected templates:

  • Data configuration required before Build phase
  • Parameter-based resolution for all connected shapes
  • Automatic population during generation
  • Ad-hoc data addition still available for extra content

Hybrid Approach

You can combine both:

  1. Configure workflow data connections for main charts
  2. Add additional data ad-hoc during Build phase
  3. Reference all data in your content
  4. AI uses both configured and ad-hoc data appropriately

Data Throughout the Workflow

Build Phase

Configured data:

  • Appears in "Connected Data (read-only)" section
  • Available for Content Assistant to reference
  • Parameters locked to configured values

Ad-hoc data:

  • Added via "Add Data Object" menu
  • Appears as editable content pieces
  • Can be modified or removed

Structure Phase

Data shapes:

  • Pre-assigned to data-connected template shapes
  • Structure AI works around these fixed assignments
  • Additional content uses remaining shapes

Populate Phase

Data population:

  • Configured data fills charts/tables directly
  • No AI interpretation of data values
  • Formatting follows template styles

Content with data references:

  • AI uses data to inform text content
  • Can summarize, analyze, compare data points
  • Creates narrative around data

Refine Phase

Data shapes:

  • Data content cannot be changed (source from collections)
  • Can adjust surrounding text and context
  • Can request different analysis or summary

Best Practices

For Recurring Reports

  1. Use workflow-based approach - Configure data connections upfront
  2. Consistent parameters - Same parameter structure each time
  3. Template shapes - Design charts/tables for your data structure
  4. Validate first - Ensure data exists before building content

For One-Off Presentations

  1. Add data ad-hoc - More flexibility during Build
  2. Let AI interpret - Use Content Assistant to analyze data
  3. Reference in prompts - Tell AI what to do with the data
  4. Multiple data sources - Mix data from different collections

For Data-Heavy Presentations

  1. Pre-configure critical charts - Main metrics via workflow
  2. Add supporting data ad-hoc - Background information
  3. Create summaries - Ask AI to distill key insights
  4. Balance automation and narrative - Data populates, AI tells the story

Common Scenarios

Monthly Report

Setup:

  • Template with monthly KPI charts (data-connected)
  • Workflow defining report structure

Flow:

  1. Start new deck from workflow
  2. Set valuation_date = current month end
  3. Validate all charts have data
  4. Build additional commentary
  5. Generate with automated chart population

Quarterly Review with Comparisons

Setup:

  • Template with comparison charts
  • Multiple data collections (current quarter, previous quarter)

Flow:

  1. Configure current quarter data for main charts
  2. Add previous quarter data ad-hoc
  3. Ask Content Assistant to create comparisons
  4. AI references both data sets in narrative

Client-Specific Presentation

Setup:

  • Template with portfolio charts
  • Portfolio holdings collection

Flow:

  1. Set portfolio_id = specific client portfolio
  2. Set valuation_date = reporting date
  3. Validate client's data exists
  4. Build custom narrative around their holdings
  5. Generate personalized presentation

Troubleshooting

"Data not appearing in chart"

Check:

  • Data configuration was validated successfully
  • Shape is properly linked to collection in template
  • Data has correct structure (columns match expected format)

"Can't add data object"

Check:

  • Data collection has files uploaded
  • Parameter values match available files
  • Resolution completed successfully before adding

"AI not using my data"

Try:

  • Reference data explicitly in prompts
  • Add data as content piece (ad-hoc)
  • Be specific about what analysis you want

"Wrong data in chart"

Check:

  • Parameters were set correctly during configuration
  • Correct file was uploaded to collection
  • No parameter override on the specific shape

Next Steps

Configuring Data for SlidesOverview