Skip to content

Export to PPTX, PDF, and Excel

Ask RITA can export query results to presentation-ready PowerPoint, PDF reports, and Excel spreadsheets — complete with charts, data tables, and branded formatting.

Table of Contents

Overview

Export Package Charts Data Table Branding
PPTX python-pptx Native PowerPoint charts Up to 15 rows Primary + secondary colors
PDF reportlab Matplotlib PNG images Up to 20 rows Primary color header
Excel xlsxwriter Native Excel charts Full dataset Branded cell formatting

All three export methods are available on SQLAgentWorkflow and return bytes that you can write to a file or stream to a client.

Installation

Export dependencies are optional. Install them as needed:

# All export dependencies
pip install python-pptx reportlab xlsxwriter matplotlib

# Or individually
pip install python-pptx    # For PPTX
pip install reportlab      # For PDF
pip install matplotlib     # For PDF charts
pip install xlsxwriter     # For Excel

Quick Start

from askrita import SQLAgentWorkflow, ConfigManager

config = ConfigManager("config.yaml")
workflow = SQLAgentWorkflow(config)

# Run a query
result = workflow.query("What are total sales by region?")

# Export to PowerPoint
pptx_bytes = workflow.export_to_pptx(result, title="Sales Report")
with open("report.pptx", "wb") as f:
    f.write(pptx_bytes)

# Export to PDF
pdf_bytes = workflow.export_to_pdf(result, title="Sales Report")
with open("report.pdf", "wb") as f:
    f.write(pdf_bytes)

# Export to Excel
excel_bytes = workflow.export_to_excel(result, title="Sales Report")
with open("report.xlsx", "wb") as f:
    f.write(excel_bytes)

ExportSettings

All export methods accept the same set of parameters, which are converted to an ExportSettings object internally:

Parameter Type Default Description
title str "Query Results" Report title (cover slide, header)
company_name str "Data Analytics" Company name for branding
include_sql bool False Include the generated SQL in the export
include_data_table bool True Include a data table (PPTX and PDF only)
chart_style str "modern" Chart style for PDF: "modern" or "classic"
brand_colors dict — Custom brand colors (PPTX and Excel only)

Brand Colors

Pass custom brand colors as RGB tuples:

pptx_bytes = workflow.export_to_pptx(
    result,
    title="Quarterly Report",
    brand_colors={
        "primary": (0, 47, 135),     # Header and accent color
        "secondary": (204, 9, 47),   # Title highlights
    }
)

Default colors:

  • Primary: (0, 47, 135) — dark blue
  • Secondary: (204, 9, 47) — red

ExportSettings Model

from askrita import ExportSettings

settings = ExportSettings(
    title="My Report",
    company_name="Acme Corp",
    include_sql=True,
    include_data_table=True,
    chart_style="modern",
    brand_primary_color=(0, 47, 135),
    brand_secondary_color=(204, 9, 47),
)

PPTX Export

PowerPoint exports create a multi-slide presentation:

Slides

Slide Content Condition
Title Report title, company name Always
Executive Summary Question, answer, SQL reasoning, visualization reasoning Always
Visualization Native PowerPoint chart from chart_data When chart data available
Data Table First 15 rows of results When include_data_table=True and results exist
SQL Query The generated SQL When include_sql=True and SQL exists
Follow-Up Questions Suggested next questions When follow-ups exist
Closing "Thank you" slide Always

Chart Types

The PPTX exporter creates native PowerPoint charts (not images). Supported chart type mappings:

  • Bar, Column, Stacked Bar/Column
  • Line, Area
  • Pie, Donut
  • Scatter
  • Multi-axis / dual-axis charts

Example

result = workflow.query("Show monthly revenue trends for 2025")

pptx_bytes = workflow.export_to_pptx(
    result,
    title="Revenue Analysis — 2025",
    company_name="Acme Corp",
    include_sql=True,
    include_data_table=True,
    brand_colors={"primary": (25, 25, 112), "secondary": (220, 20, 60)},
)

with open("revenue_2025.pptx", "wb") as f:
    f.write(pptx_bytes)

PDF Export

PDF exports create an A4 document with sections:

Sections

Section Content Condition
Title Block Report title, company name, date Always
Query Analysis Question and answer Always
SQL Query The generated SQL When include_sql=True and SQL exists
Chart Matplotlib-rendered chart as PNG When chart data available
Data Table First 20 rows of results When include_data_table=True and results are dicts
Follow-Up Questions Suggested next questions When follow-ups exist

Chart Styles

Style Description
"modern" Clean, minimal style with modern colors
"classic" Traditional chart styling

Example

result = workflow.query("Customer distribution by state")

pdf_bytes = workflow.export_to_pdf(
    result,
    title="Customer Geography Report",
    chart_style="modern",
    include_sql=False,
)

with open("customers_by_state.pdf", "wb") as f:
    f.write(pdf_bytes)

Excel Export

Excel exports create a workbook with data and charts:

Sheets

Sheet Content Condition
Data Title, question, answer, optional SQL, full results table, chart Always
Summary Follow-up questions When follow-ups exist

Features

  • Full dataset — Unlike PPTX (15 rows) and PDF (20 rows), Excel includes all result rows
  • Native Excel charts — Bar, column, line, pie, and multi-axis chart types
  • Formatted headers — Branded colors on header row and title cells
  • Chart-aware columns — When chart_data exists, column headers are derived from the chart metadata

Example

result = workflow.query("All orders from the last 30 days with customer details")

excel_bytes = workflow.export_to_excel(
    result,
    title="Recent Orders Export",
    company_name="Acme Corp",
    include_sql=True,
    brand_colors={"primary": (0, 100, 0)},
)

with open("recent_orders.xlsx", "wb") as f:
    f.write(excel_bytes)

API Reference

SQLAgentWorkflow.export_to_pptx()

def export_to_pptx(
    self,
    output_state: WorkflowState,
    title: str = "Query Results",
    company_name: str = "Data Analytics",
    include_sql: bool = False,
    include_data_table: bool = True,
    chart_style: str = "modern",
    brand_colors: Dict[str, tuple] = None,
) -> bytes:
    """
    Export query results to a PowerPoint presentation.

    Raises:
        ImportError: If python-pptx is not installed
        ExportError: If export fails
    """

SQLAgentWorkflow.export_to_pdf()

def export_to_pdf(
    self,
    output_state: WorkflowState,
    title: str = "Query Results",
    company_name: str = "Data Analytics",
    include_sql: bool = False,
    include_data_table: bool = True,
    chart_style: str = "modern",
) -> bytes:
    """
    Export query results to a PDF document.

    Raises:
        ImportError: If reportlab is not installed
        ExportError: If export fails
    """

SQLAgentWorkflow.export_to_excel()

def export_to_excel(
    self,
    output_state: WorkflowState,
    title: str = "Query Results",
    company_name: str = "Data Analytics",
    include_sql: bool = False,
    include_data_table: bool = True,
    chart_style: str = "modern",
    brand_colors: Dict[str, tuple] = None,
) -> bytes:
    """
    Export query results to an Excel workbook.

    Raises:
        ExportError: If xlsxwriter is not installed or export fails
    """

Lower-Level Functions

These are available for advanced use cases:

from askrita.sqlagent.exporters.core import create_pptx_export, create_pdf_export
from askrita.sqlagent.exporters.excel_exporter import create_excel_export
from askrita.sqlagent.exporters.chart_generator import (
    generate_chart_bytes,        # PNG bytes via matplotlib
    get_chart_data_for_export,   # Extract chart data from WorkflowState
    add_native_pptx_chart,       # Add native chart to a PPTX slide
)

Troubleshooting

ImportError: Missing Export Dependencies

Symptom: ImportError: python-pptx is required for PPTX export

Install the required package:

pip install python-pptx   # For PPTX
pip install reportlab      # For PDF
pip install xlsxwriter     # For Excel
pip install matplotlib     # For PDF charts

Empty Charts in Export

Symptom: Export has no chart, only data table.

  • Ensure the query result has chart_data (enable the choose_and_format_visualization workflow step)
  • Check that visualization is not empty in the result
  • Some queries return data that is not suitable for charting

ExportError

Symptom: ExportError during export.

  • Check that the output_state contains valid data (results, answer)
  • Use --verbose when querying to see if the workflow completed successfully
  • Verify chart data is well-formed (non-empty labels and values)

PDF Charts Missing

Symptom: PDF exports have no chart even when chart data exists.

  • Install matplotlib: pip install matplotlib
  • The PDF exporter uses matplotlib to render charts as PNG images

See also: