AI didn't make customer intelligence obsolete. It made it essential.
Everyone on your team is already using AI. PMs have Claude open. Engineers are in Cursor. Someone's got a ChatGPT tab they never close. Designers are prototyping in Figma Make.
These tools are genuinely great. But they all have the same problem: they don't have your customer data.
When a PM asks ChatGPT to summarize what customers think about a feature, it's working with whatever got pasted into the prompt. It doesn't know who said it. It can't tell the difference between a churning enterprise account and a happy free user. It can't trace anything back to a real conversation. And whatever it produces lives and dies in that one chat window—nobody else on the team ever sees it.
That's the gap. General-purpose AI is great at reasoning. It's terrible at knowing your customers. And that's exactly why we've built Dovetail to work with your AI stack, not against it.
Dovetail now connects to the tools your team already uses
We've shipped connectors for ChatGPT, Claude, Cursor, Figma Make, and any tool that supports MCP (Model Context Protocol).
In practice, this means your AI tools can now pull directly from your Dovetail workspace. Full context. Proper attribution. Your complete history of customer conversations.
A PM can ask ChatGPT a question about customer feedback and get an answer grounded in actual data from Dovetail—not a hallucinated guess. An engineer in Cursor can check what customers have been saying about a feature before writing a single line of code or prompting. A designer in Figma Make can pipe real user quotes into a prototype so the output reflects what people actually said.
The intelligence stays in Dovetail. Your other tools become more effective.
Better inputs beat better prompts
Everyone says, “write better prompts.” That’s fine, but it only gets you so far when the input is a pasted excerpt with zero context.
The real unlock is better inputs. Structured, attributed, continuously updated customer data. That's what Dovetail gives your AI tools.
When Dovetail is connected, the answers are fundamentally different. Claims trace back to real people. Feedback filters by segment, ARR, plan tier. Insights are grounded in evidence, not inferred from thin context. And the same intelligence is available to everyone—not locked in one person's ChatGPT chat history.
This is the difference between asking AI to guess and asking it to look things up in a system of record.
What you can do today
ChatGPT—Dovetail is available as a native app inside ChatGPT. Connect your workspace, and ChatGPT can search and reference your customer data directly. No setup beyond logging in.
Claude—Connect Dovetail to Claude Desktop or Claude Code via our MCP server. Your workspace becomes a live data source that Claude can query.
Cursor—Developers can connect Dovetail's MCP server to Cursor, bringing customer intelligence into their coding environment. Before building a feature, ask what customers actually said about it.
Figma Make—Pull real user quotes, feedback, and synthesized research insights from Dovetail into your prototypes. Design from what users actually want, not what you think they want.
Anything that supports MCP—Dovetail's hosted endpoint works with any MCP-compatible tool—Windsurf, Microsoft Copilot Studio, and more. If your tool supports MCP, it can connect to Dovetail.
“Why do we need Dovetail if we have Claude?”
Fair question. We use Claude too, and in practice, these tools do very different jobs.
Claude is an inference engine. Brilliant at reasoning, summarizing, and generating. But it has no data. It doesn't know your customers. It can't segment feedback by revenue. It can't share what it finds with the rest of your org or support collaborative sessions.
Dovetail is the intelligence layer that sits beneath those tools. It connects to the places where those conversations already happen—like Gong, Zendesk, Intercom, Salesforce, Slack, app reviews, and surveys—and makes that data structured, searchable, and attributable. Every insight links back to a specific quote from a specific person at a specific company. Every analysis can be filtered by the dimensions that actually matter to your business.
With our connectors, these become complementary. Dovetail brings the data and attribution. Your AI tools bring the reasoning. The result is customer intelligence that's powerful and trustworthy—available to the whole team, in the tools they already use.
A folder full of feedback is not a strategy
From here, a tempting idea might follow: if AI can read documents, why not just point it at a shared folder and call it a day? Skip the platform entirely, dump everything into Google Drive or SharePoint, point Claude at it, and call it a day. It doesn't work. Your Drive doesn't have your customer data—it has whatever someone remembered to upload. Feedback lives across dozens of tools, owned by different teams, in different formats. Even if an eng team built a pipeline to automate that, they'd only have solved the easiest part: getting raw files into a folder.
An LLM can summarise a transcript, but it can't tell you that the same pain point appeared 47 times across EMEA enterprise customers in your pharma vertical this quarter. That requires structured, enriched data—not a file search. And the way data gets served to an LLM matters. Dovetail's MCP is built to return structured customer intelligence—highlights, attributed quotes, themes, sentiment—in a way an LLM can actually use. Drive's MCP returns documents. It handles transcripts, tickets, survey responses, videos, and reviews all the same way: as files. We handle them as customer data, because that's what they are.
And when your VP asks, “How do we know customers want this?” citing a Drive filename isn't the same as tracing a claim to a specific speaker, a specific moment, connected to every other instance of that pattern across your workspace. What sounds like a simple shortcut is actually a proposal to build a customer intelligence platform from scratch using horizontal tools—and most teams discover they've spent months on the plumbing and still don't have the intelligence.
This is a team sport
When someone pastes transcripts into ChatGPT, the output is private. It lives in their chat history. Nobody else sees it, builds on it, or verifies it. The next person who needs the same answer starts from scratch.
Dovetail makes that intelligence shared and persistent. It compounds and gets richer over time. Research from last quarter is still searchable. A support trend in Channels is visible to leadership, and other teams can access and act on that same intelligence.
The connectors extend this. When anyone on your team queries Dovetail through ChatGPT or Claude, they're drawing from the same shared intelligence, not their own private collection of pasted excerpts.
Your customer data deserves more than a chat window
If you're making product decisions based on customer feedback—and you should be—then where that data lives matters. How it's attributed matters. Who can access it matters. Whether it meets your compliance requirements matters.
Dovetail handles PII controls, automatic redaction, SOC 2, GDPR, and EU AI Act compliance. For teams in pharma, healthcare, and finance, this isn't optional. It's the difference between being able to use AI for customer intelligence and not being able to at all.
General AI is getting better every day. And that's good for us, because that makes the data layer more important, not less. The better your AI tools get at reasoning, the more valuable it is to give them the right data to reason over.
Dovetail is that layer. And now it connects to everything.
Related Articles
