Coming soon · Early access
Your analysis lives in QualIntel. Your evidence follows you into the AI you write with.
The QualIntel Connector brings your researcher-accepted evidence into Claude, ChatGPT, or Microsoft Copilot — with every retrieval logged to your project’s audit trail and summarised in your AI disclosure statement. Not another chatbot. A logged, defensible bridge between your coded data and the place you actually write.
How it works
Do the rigorous work in QualIntel
Upload your research design, confirm your codebook, and accept or reject every piece of candidate evidence. Interpretation stays with you — that part never moves.
Connect your AI assistant
Add QualIntel as a connector in Claude, ChatGPT, or Microsoft Copilot and sign in with your QualIntel account. No files to export, no transcripts to paste.
Write with your evidence beside you
Ask for the strongest accepted excerpts for a theme, with participant IDs and source provenance — while you draft your chapter. Every retrieval is logged to your project's audit trail.
What it deliberately won’t do
Most AI integrations compete on what they can do. This one is defined by what it refuses to do — because that is what keeps your analysis defensible.
It will not find your themes
There is no tool for “analyse my data” — by design. The connector retrieves evidence you have already reviewed and accepted. Researcher-led interpretation is the point, not a limitation.
It will not serve unreviewed suggestions
Only evidence you accepted in QualIntel is available through the connector. If a project hasn't completed evidence review, the connector says so and points you back to the workflow.
It will not create an invisible workflow
Every retrieval — which tool, what query, which excerpts were returned, when — is written to your project's audit trail, and summarised in your AI disclosure statement.
The sentence your methods section gets
When the connector ships, your AI disclosure statement gains a clause no other workflow can honestly print:
“Evidence was retrieved into external AI writing tools via QualIntel’s logged connector; the full retrieval log — tool, query, excerpts returned, and timestamps — is available in the project audit trail.”
To be precise about the boundary: QualIntel logs every retrieval with provenance. It does not capture the conversation inside your AI assistant. A logged retrieval record is not a transcript of your chats — and it is categorically more defensible than pasting data into a chatbot with no record at all. Read more on the qualitative research audit trail and the AI disclosure statement.
Frequently asked questions
What is the QualIntel Connector?
The QualIntel Connector links your QualIntel OS projects to AI assistants that support connectors — Claude, ChatGPT, and Microsoft Copilot. It lets you retrieve your researcher-accepted evidence, codebook, and audit trail from inside the AI tool you already write with. It is built on the open Model Context Protocol (MCP) standard and is currently in development, with an early-access waitlist open.
Will the connector analyse my data in ChatGPT or Claude?
No — deliberately. The connector performs AI-assisted evidence retrieval only: it returns excerpts you have already reviewed and accepted inside QualIntel OS, with participant IDs and source provenance. It has no tools for generating themes or findings. Interpretation remains researcher-led, inside the audited QualIntel workflow.
Is my AI use through the connector auditable?
Every evidence retrieval made through the connector is logged with provenance — the tool used, the query, the excerpts returned, and the timestamp — into your project's audit trail, and summarised in your AI disclosure statement. The conversation you have inside Claude or ChatGPT itself is not captured by QualIntel, so the connector gives you a logged retrieval record, not a transcript of your AI chats. That is still a categorical improvement over pasting transcripts into a chatbot with no record at all.
Which AI apps will the connector support?
Any AI assistant that supports remote connectors via the Model Context Protocol (MCP), which today includes Claude (web and desktop), ChatGPT, and Microsoft Copilot, with more clients adding support. You connect once by signing in with your QualIntel account — no API keys to manage.
Do I need a QualIntel OS subscription to use the connector?
The connector will be included in the Researcher plan. Free accounts will be able to try it against the built-in demo project to see how logged evidence retrieval works before subscribing.
Why does this matter for my methods section?
Examiners and supervisors increasingly ask how AI was used in qualitative analysis. If you paste transcripts into a chatbot, there is no record to point to. With the connector, you can state that all AI access to your evidence went through a logged channel, cite the retrieval log in your audit trail, and include the summary in your AI disclosure statement.
The connector serves projects with accepted evidence — so the best way to be ready is to start coding now. Free, no time limit.
Start freeRelated: Integrations · AI disclosure statement · Research audit trail · Pricing