Extracting project dates from messy Gantt charts with AI
The goal is to pull from PPTX and PDF project packs, including activities, milestones, risks, issues, owners, statuses, and dates. The hardest part is the .
These charts can look very different from one document to another, with months, quarters, years, week-start labels, fiscal periods, mixed time scales, different bar styles, milestone icons, legends, and layouts. Some files contain real PPTX shapes, while others contain or flattened PDFs that are harder to inspect directly.
A reliable system would likely need a mix of LLM and VLM reasoning, rule-based extraction, OCR, file parsing, and coordinate-based date mapping. The key challenge is getting start dates, end dates, and milestone dates from many chart formats without each document style.
Key points
- The target files are PPTX and PDF project packs such as status reports and planning decks.
- The data to extract includes activities, milestones, risks, issues, owners, statuses, and dates.
- are difficult because their time scales, labels, bars, icons, and layouts vary widely.
- Some charts are editable PPTX shapes, while others are or flattened PDFs.
- A practical approach would combine LLM, VLM, OCR, parsing, and coordinate-based date mapping.