From Audit Findings to Management Letters in Minutes with AI

This case study showcases how an AI-powered Management Letter tool transformed a frustrating, time-consuming task into a fast, reliable, and hassle-free process. The product is part of a larger ecosystem aimed at reshaping how auditors conduct audits and communicate with clients efficiently.

Allison, a fictional yet relatable auditor — once juggled scattered audit notes, tedious evidence checks, and manual source tracing. Today, a custom-built AI tool scans thousands of pages, extracts key issues, and drafts Management Letters with traceable citations and properly formatted sections.

Built as a Microsoft Word add-in, it cut preparation time by 80%, reduced errors by 70%, and doubled issue-identification speed — clear proof that well-designed AI solutions can solve real professional challenges while boosting productivity.

• 80%
  faster draft prep time
• 70%
  fewer citation errors
• 2×
  issue-identification speed

The tool is slated to be released firmwide by Q2 2026, benefiting both internal and external auditors. The rollout will require domain-specific audit data and collaboration with subject matter experts to fine-tune issue identification and citation generation.

"This is a must have for the global team, lets bring this up to the global heads to have a discussion how could we introduce the new way of management letter preparation. The intuitiveness of the management letter preparation experience is worth looking into." — Audit Innovation Partner

Meet Allison

Imagine Allison staring at an empty Word document, the cursor blinking like a metronome for billable hours. "Where do I even begin pulling all these findings together?" she sighs.

And she’s not alone. Across audit firms, Management Letter prep means chasing down notes, evidence, and comments scattered across platforms. Even after fieldwork’s done, the real grind begins — turning findings into a clean, client-ready document.

The Challenge

“The ask is simple — trace every finding back to its source,” Allison says.

But with larger, more complex engagements, this means becoming a human search engine, hunting for the right file, page, and paragraph. What should be a straightforward job turns into an endless folder search. It’s frustrating, inefficient, and prone to errors.

How We Retought the Process

Our product team turned to generative AI to fix it. But early chats revealed a gap — while most auditors had heard of tools like ChatGPT, they didn’t know what AI could do for Management Letters. We shifted our design approach. Instead of asking what they needed, we showed them what was possible.

Using a “reverse double diamond” framework, we quickly delivered prototypes and refined through user feedback:

1. Mapped current Management Letter drafting workflows
2. Pinpointed friction and pain points
3. Explored how AI/ML could support key tasks
4. Embedded AI actions naturally into their tools

By co-creating with a small pilot group, we built features they hadn’t asked for but quickly came to rely on.

AI Assisted Management Letter Drafting

The tool uses advanced natural language processing to scan audit evidence, extract issues, and link findings to their source documents. It’s more than a search tool — it’s a contextual knowledge engine built for audits.

For Allison, here’s how it works:

1. She inputs her fieldwork findings
2. The AI suggests well-written, grouped findings for her review
3. Each suggestion includes source links and context
4. With one click, she inserts them into her Management Letter draft
5. The add-in auto-formats and ensures compliance with firm standards