Blog
Inside Seeqle

Seeqle comes to your AI assistant with the AI connector (MCP)

By
Jean Eudes Yahouedeou
5min
Published on
10/11/2026 ·
Updated on
10/11/2026
Share this post
Seeqle comes to your AI assistant with the AI connector (MCP)
Share this post

Since November 9, 2026, you can run Seeqle from your AI assistant. The AI connector (MCP) links Seeqle to Claude, ChatGPT, Gemini and Perplexity: you ask in plain language, and the assistant works in Seeqle with your permissions. Here is why we built it, what it changes for a recruiter, and what stays in your hands.

Why bring Seeqle into your AI assistant?

More and more recruiters keep their AI assistant open all day. They draft there, summarize there, prepare there. Then they switch tabs, open another tool, and start clicking again.

I don’t believe in one more screen. HR software that makes you come to it adds friction for teams that already have too much. With the AI connector, Seeqle goes the other way: it comes to where you already work.

It is the natural next step of what we have been building from the start. Our agents execute, they don’t just assist. One thing was missing: being able to talk to them without opening Seeqle.

What is the AI connector (MCP), in practice?

MCP, short for Model Context Protocol, is the open standard that lets AI assistants plug into software. Seeqle’s AI connector follows this standard. Once connected, your assistant can work in Seeqle when you ask it to.

It works with four assistants:

  • Claude;
  • ChatGPT;
  • Gemini;
  • Perplexity, on its paid plan.

The connector has been available since November 9, 2026. In Seeqle, you will find it under Settings › AI connector (MCP).

What can you ask your assistant?

What used to take several screens now fits in one sentence. The connector covers six uses.

UseExample request
Create a job offer“Create a permanent maintenance technician job in Lyon from this job description.”
Create a form“Prepare a short application form for this job, with two questions about availability.”
Prepare a campaign“Prepare a campaign for this job. I’ll review it before submitting.”
Generate the visuals“Generate the visuals for this campaign.”
Track results and reports“How are my campaigns doing this month? Summarize the latest report.”
Find your candidates“Show me the best-rated candidates for the technician job.”

Take an ordinary week. On Monday morning, one question gives you the status of your campaigns, without opening a dashboard. On Tuesday, a manager sends you a job description: you paste it into your assistant, which creates the job offer and prepares the campaign. On Friday, before your interviews, you ask it for the best-rated candidates. Three requests, zero navigation.

What happens to the Attract Agent and the Match Agent?

Seeqle’s two agents don’t change. What changes is the way in.

The Attract Agent reads the job offer, picks the channels, sets the targeting, writes the ad copy and generates the visuals. The Match Agent analyzes each application and gives it a score out of 5, with the reasons. The decision stays human.

Until now, you had to go to them. From now on, one sentence in your assistant is enough to call on Attract or find the candidates scored by Match. To me, that is what an AI agent should be: a colleague you call on wherever you are, not software you have to learn to operate.

Four-step diagram: you ask in plain language, the assistant works in Seeqle with your permissions, Attract prepares the campaign and you find the candidates scored by Match, you decide to submit the campaign

What safeguards protect your campaigns and your candidates?

An assistant that acts fast has to act within clear limits. We set three rules from day one:

  • Your permissions, your workspaces. The assistant works with the user’s permissions, in the user’s workspaces. It has no more power than you do.
  • Drafts until you give the green light. Campaigns prepared by the assistant stay drafts until you submit them. The decision to launch stays yours.
  • Protected candidates. Candidates are read-only and anonymized. The assistant helps you find them, it does not change them.

These rules add to the ones we apply across Seeqle, detailed on our security and data page.

How do you connect Seeqle to your assistant?

  1. In Seeqle, open Settings › AI connector (MCP).
  2. Choose your assistant: Claude, ChatGPT, Gemini or Perplexity (paid plan).
  3. Follow the instructions shown for that assistant.
  4. Make your first request.

If you are new to AI agents in recruiting, our checklist for choosing an agentic AI HR tool helps you ask the right questions, and our guide to AI in recruitment gives the context.

What comes next?

This launch is a beginning. We look at what you ask, what works the first time and what gets stuck. That is what will shape what comes next, not a roadmap written in advance.

My conviction is simple: HR software will increasingly be run in natural language. Recruiters will pick their assistant, and they will expect their tools to plug into it. Seeqle made that choice, because that is where recruiters already work.

Frequently asked questions

Which AI assistants work with Seeqle’s AI connector?

Seeqle’s AI connector (MCP) works with Claude, ChatGPT, Gemini and Perplexity, on its paid plan. It follows MCP, the open standard that lets AI assistants plug into software. It has been available since November 9, 2026, under Settings › AI connector (MCP) in Seeqle.

Can my assistant launch a campaign without my approval?

No. Campaigns prepared by your assistant stay drafts until you submit them. The assistant can create the job offer, prepare the campaign and generate the visuals, but the decision to submit the campaign is always yours. Nothing goes out without your green light.

Can the assistant see candidates’ personal data?

Not their identity: candidates are read-only and anonymized. Your assistant can help you find a job’s candidates and spot the best-rated ones, without changing their records. It works with your permissions, in your workspaces, and never has more power than you do in Seeqle.

What is the MCP protocol?

MCP, short for Model Context Protocol, is the open standard that lets AI assistants plug into software. It is the standard Seeqle’s AI connector follows to connect to Claude, ChatGPT, Gemini and Perplexity. In Seeqle, you will find it under Settings › AI connector (MCP).

Model Context Protocol, official documentation: modelcontextprotocol.io

The next step

How do you spend less time sorting applications?

The Match Agent scores every application as it arrives, and you decide.

Try it free

No credit card. 14 days free.

Further reading

Get applications that are already sorted

Your jobs shown where the talent is, every application scored as it arrives.

Request a demo
Try it free

No credit card. 14 days and €100 of ad distribution free.

Award-winning
Selected among
The AI startups of the year
France Digitale
Selected among
The 100 start-ups to invest in
Challenges
HAVAS, Station F
Startup Accelerator
Républik RH
Gold Medal, Recruitment category
Meta, Station F
Startup Garage Paris
Copied!