PowerPoint Report Automation Engine

Problem

Analysts were spending hours manually assembling PowerPoint reports — copy-pasting charts, updating placeholders, and fixing formatting drift across slide masters.

Business context

A regulated reporting workflow where the same deck structure was reproduced weekly and monthly for executive audiences, with strict branding and content rules.

My role

Engineer-owner of the .NET automation engine — template format, slide composition, queueing, and the operator UI used by the analyst team.

Solution

• Defined a template format with named placeholders for text, tables, charts, and images.
• Built a composition engine that hydrates placeholders from structured inputs and produces a clean PPTX.
• Queued generation runs so a single request can fan out into dozens of decks.
• Surfaced status, errors, and downloadable artifacts in a simple operator dashboard.

Architecture highlights

• Open XML SDK for safe, deterministic PPTX manipulation.
• .NET Core background workers consuming a durable queue.
• Inputs validated against a typed schema before any slide is touched.
• Artifacts written to object storage with retention metadata.

Challenges

• PPTX is unforgiving — a single malformed shape can corrupt the whole deck.
• Templates drift over time; needed validation that catches breakage before generation.
• Charts had to remain editable inside PowerPoint, not flattened images.

Outcome / impact

• Reduced manual effort for recurring decks.
• Improved consistency across branded reports.
• Enabled faster reporting cycles for analysts.

What I learned

• Schema-first input is the single biggest reliability win for document automation.
• Validation belongs upstream — generation should assume clean input or fail loudly.