I used to spend 4+ hours synthesizing user interviews.

Highlighting quotes. Tagging themes. Building a spreadsheet. Writing the summary deck.

Last month I tried something different.

Uploaded 12 interview transcripts to Claude and asked:

"Identify the top 5 recurring pain points across these interviews. Include direct quotes as evidence. Flag any contradictions between users."

8 minutes. Done.

But here's where it got interesting.

I didn't just get a wall of text back. I asked Claude to build an Artifact — a visual research dashboard that updates as I paste new interviews.

Pain points on the left. Quote evidence on the right. Contradiction flags highlighted in red.

A reusable tool. Not a one-time answer.

The prompts that unlock this:

"Build an interactive research synthesizer. Let me paste interview transcripts and show me: top pain points, supporting quotes, user segments, and contradictions."

"Create a user persona card based on these 5 interviews. Include goals, frustrations, and actual quotes."

"Make a feature prioritization matrix from this research. Map user pain points to potential solutions and rank by frequency."

Each one generates a working tool I can use again tomorrow.

The shift:

Synthesis used to be the bottleneck between research and action.

Now the bottleneck is asking the right questions.

Most PMs are still treating AI like a search bar. Type question, get answer, close tab.

But when you start building with it — templates, dashboards, synthesizers — the leverage compounds.

I put together a deep dive on Claude Projects & Artifacts with 60+ templates you can copy-paste (research synthesizers, RICE calculators, persona generators, the whole toolkit):

No fluff. Just the prompts and setups that actually work.