Transforming Presentation Decks into Actionable Knowledge Bases, According to Yonathan Cohen

In a recent LinkedIn post, Yonathan Cohen explores a novel approach to leveraging presentation decks, transforming them from static documents into dynamic, queryable knowledge bases. Cohen argues that many companies fail to fully utilize the valuable information contained within their slide decks, allowing them to languish in forgotten folders.

Cohen highlights the inherent value proposition of these decks:

“Your Gamma decks are your institutional knowledge. Already distilled, hierarchized, human-curated.”

However, he points out a critical flaw in their typical storage and accessibility. As Yonathan Cohen notes, these valuable resources often become inaccessible:

“But they die in a folder nobody reopens.”

From Static Storage to Interactive Knowledge

Yonathan Cohen proposes a method to overcome this limitation by integrating AI, specifically mentioning Claude, to interact with presentation libraries. This allows users to query their decks in a conversational manner, extracting specific information or even regenerating content based on updated data.

Querying Institutional Knowledge

Cohen outlines several examples of the types of questions that can be posed to an AI-powered knowledge base derived from presentation decks. This demonstrates the practical applications for sales teams, marketing departments, and leadership.

• “What did we promise in last quarter’s pitch?”
• “Find the 3 strongest case studies we’ve ever shipped.”
• “Pull up our latest investor deck.”
• “Regenerate the investor deck with our new ARR numbers.”
• “What was our angle when we pitched HubSpot?”

This capability, according to Yonathan Cohen, moves beyond simple document retrieval to intelligent synthesis and repurposing of information.

The Process of AI-Powered Deck Interaction

Yonathan Cohen further elaborates on the underlying mechanism that enables this transformation. The process involves the AI system first browsing the entire library of decks, then reading relevant documents in detail, and finally providing answers with cited sources or generating new content.

Cohen explains the workflow:

Step 1 → Claude browses your full deck library.

Step 2 → Claude reads the relevant decks in full.

Step 3 → Claude answers with cited sources, or regenerates a deck.

This structured approach, as described by Yonathan Cohen, ensures that the AI can effectively understand and utilize the content within the presentations. He emphasizes that this method allows for the continuous activation of institutional knowledge, preventing it from becoming obsolete.

By shifting from passive storage to active querying, Yonathan Cohen suggests that businesses can unlock significant value from their existing presentation assets, making historical data and strategic communications readily accessible and actionable.