DeckForge – PPT Template Reformatter

Delivered scope

• Transforms raw source slide decks into on-brand PowerPoint presentations automatically.
• Auto-detects template image and text placeholders directly from PPTX files without coordinate entry.
• Subdivides image areas into side-by-side, 2x2, or N-cell grids with center-crop alignment.
• Applies automatic font shrink-to-fit to guarantee clean typography without text truncation.

Technology

System approach

• Template Engine: python-pptx parser auto-detecting shape geometry, dimensions, and text styling without manual coordinate configuration.
• Grid Layout Core: Custom grid engine calculating side-by-side, 2x2, and N-cell subdivisions with center-crop Pillow transformations.
• Text Normalization: Auto-scaling typography engine adjusting font size proportionally to fit strict placeholder bounds without truncation.
• Web Console & Preview: Flask web service featuring live template library management, slide previews, and PDF export pipelines.

Challenges

• Parsing heterogeneous PowerPoint shapes, groups, and text frames across diverse legacy and modern office versions without corrupting slide XML.
• Maintaining visually balanced grid layouts when source slides contain variable numbers of images (1, 2, 3, 4, or more).
• Preventing text overflow or clipping on small placeholders without truncating essential slide copy.

Solutions

• Implemented placeholder detection heuristics that identify target frames by type, hierarchy, and bounding-box area directly from PPTX OpenXML.
• Engineered an adaptive grid subdivision algorithm that splits image frames into symmetrical rows/columns with aspect-ratio preserving center-crops.
• Built a font shrink-to-fit routine that iteratively reduces font size and line spacing to achieve zero-overflow text placement.

Key outcomes