Mischa van den Burg

A business owner with ~400 call transcripts (sales and Q&A) wants to use AI to extract insights like top client problems and purchase objections, but recognizes that naively dumping all transcripts into a single Claude project won't scale. They're seeking architectural advice on how to approach this analysis problem effectively.

If you’re smart and know about AI I need your thoughts on this.

I have around 200 transcripts of sales calls, and 200 transcripts of Q&A calls that I did with my clients.

I want to analyze these to get insight into certain questions, like "what are the top 5 problems my clients have" or "what is the #1 reason why these clients didn't buy"

Now obviously, I cannot pump all of this into a Claude project and ask it questions, we need a better approach.

How would you do it?

<p>If you’re smart and know about AI I need your thoughts on this.

I have around 200 transcripts of sales calls, and 200 transcripts of Q&amp;A calls that I did with my clients.

I want to analyze these to get insight into certain questions, like &quot;what are the top 5 problems my clients have&quot; or &quot;what is the #1 reason why these clients didn't buy&quot;

Now obviously, I cannot pump all of this into a Claude project and ask it questions, we need a better approach.

How would you do it?</p>