Beatview MCP server for recruiting

Run your hiring from Claude, ChatGPT or any AI agent.

Connect Beatview once. Then open roles, check who applied, compare the shortlist and send interview invitations just by asking. Beatview has already screened and interviewed the candidates — your assistant brings back the evidence, and you make the call.

Serverhttps://platform.beatview.ai/mcp

Works with Claude, ChatGPT, Gemini, Cursor and other MCP-compatible tools · included on every plan

What it is

The Beatview MCP server is a remote Model Context Protocol endpoint that connects Claude, ChatGPT, Gemini, Cursor and other MCP-compatible AI tools to a Beatview account, on every Beatview plan. Once a recruiter approves the connection, the assistant can create and publish jobs, list and compare candidates with their AI resume and interview scores, show the hiring pipeline, move candidates between stages and send AI interview invitations — and it confirms with the recruiter before moving a candidate or sending an invitation.

Why it is different

Your assistant doesn’t start from a pile of CVs.

Connecting an assistant to a database gives it records. Connecting it to Beatview gives it candidates your agent has already screened, interviewed and ranked — so every answer comes with evidence.

Already screened

Every application has been scored against your role criteria, with the resume evidence and concerns attached.

Already interviewed

Candidates who passed your rule have answered structured video questions, assessed one answer at a time.

Already ranked

Completed interviews are scored and ranked, so “who are the strongest?” has a real answer.

Yours to decide

The assistant brings back the evidence and a link to the record. Moving someone forward is your call.
A week on one role

Ask when you want. The agent keeps working when you don’t.

Screening, stage rules and interview invitations run inside Beatview around the clock. The chat is where you check in and decide.

  1. Step 01

    Monday · you ask

    “Open a Senior Account Manager role in London and publish it.” The job is live on your career page, with a link to post anywhere.
  2. Step 02

    All week · Beatview works

    Every application is screened as it arrives. Your stage rule invites the ones above your threshold to the AI interview — no chat needed.
  3. Step 03

    Every morning · a quick check

    “Who’s new since yesterday?” Ask it yourself, or schedule the prompt if your assistant supports recurring tasks.
  4. Step 04

    Thursday · you compare

    “Compare the top three.” A comparison table, strongest first, with the answers behind each score one click away.
  5. Step 05

    Friday · you decide

    “Move Aisha to face-to-face.” The assistant asks you to confirm, Beatview does it, and the action is logged.
What your assistant can do

Jobs, candidates and the pipeline — all from the chat.

These are the actions the server exposes today. Two of them change a candidate’s journey, so the assistant asks you first.

Actions available through the Beatview MCP server
ActionWhat it doesAsk it like this
create_jobCreates a job as a draft or published“Open a Senior React role in Dubai, remote.”
update_jobUpdates title, location, description and more“Change the Account Manager role to hybrid.”
publish_job / unpublish_jobMakes a job public on your career page, or hides it“Take the Warehouse Lead role offline.”
get_job_linkReturns the shareable link plus a ready-to-send message“Give me the link and a LinkedIn post for this role.”
list_jobsLists jobs with status, applicants and links“Which roles are live and how many applied to each?”
get_career_pageReturns the career page link and published job count“What’s our careers page link?”
list_candidatesLists candidates for a job with stage, location and AI scores“Who are the strongest applicants for this role?”
search_candidatesFinds candidates across all jobs by name, email, location or score“Anyone in Manchester with an interview score above 80?”
get_candidateFull profile: screening answers, resume insights, AI scores, interview results, work style, prep report and notes“Brief me on Aisha before my call.”
get_interviewInterview evaluation: summary, per-dimension scores and each question with the AI assessment and a Beatview link“How did she answer the negotiation question?”
get_pipelineShows every stage of a job, how many candidates sit in each and who they are“Show me the pipeline.”
move_candidate_to_stageAsks you firstMoves a candidate to another stage, using the same rules as the app“Move Aisha to face-to-face.”
invite_candidate_to_interviewAsks you firstSends the branded AI interview invitation email“Send Daniel the interview invite.”
Prompt library

Copy a prompt. Paste it into your assistant.

Beatview also offers starter prompts inside connected assistants — “Show me the pipeline”, “Compare the top candidates”, “Who needs action today” — so nobody has to learn a new way of asking.

Open and share roles

Daily triage

Shortlist and decide

Client and team updates

Answers you can read

Beatview tells the assistant how a recruiter wants to see it.

Instead of handing over raw data, the server sends layout guidance with each result, so Claude, ChatGPT and others show it the same way every time.

Pipeline → kanban

One column per stage with counts in the headers, compact candidate cards and a line on where the bottleneck is.

Numbers → charts

Stage distribution, score spread and interviews completed shown as a chart, not a wall of figures.

One candidate → profile card

Scores and location, strengths and concerns side by side, the interview breakdown, then your notes.

Several → comparison table

Strongest first, with the differences that decide it highlighted — and a Beatview link on every row.
Connect in a minute

One address. One approval screen.

No API keys to paste and nothing to install. Your assistant sends you to Beatview, you approve the connection, and you can see or revoke it any time.

Server address

https://platform.beatview.ai/mcp

Configuration file

{
  "mcpServers": {
    "beatview": {
      "url": "https://platform.beatview.ai/mcp"
    }
  }
}
Read the full setup guide
  1. 01

    Step 01

    Add the address

    In Claude, ChatGPT, Gemini, Cursor or another MCP-compatible tool, add a custom connector or custom app with the Beatview server address.
  2. 02

    Step 02

    Press Connect

    Your assistant opens Beatview. Sign in if you are asked to.
  3. 03

    Step 03

    Review the request

    Beatview shows which agent is asking for access and what it will be allowed to do.
  4. 04

    Step 04

    Allow access

    Click Allow access and you are connected. Disconnect it any time from Integrations → MCP Server → Settings.
Control

Connected, not handed over.

Candidate data is sensitive, and an agent that can act needs limits. These are built into the server.

AI scores are always presented as signals to review, never as a hiring decision. A person decides who moves forward.

  • Every connection is tied to one Beatview account — an agent can never see another company’s data.
  • A connection needs you to be signed in and to approve the agent explicitly.
  • The assistant is told to confirm with you before it moves a candidate or sends an invitation.
  • Interview recordings are never sent to the agent. Each answer comes with a Beatview link that only signed-in members of your account can open.
  • Resume files come back as a temporary download link that expires after an hour.
  • Each connection is capped at 60 calls a minute and 5,000 a day, so a runaway agent cannot hammer your account.
  • Every action is logged in the MCP Server settings, and you can revoke a connection instantly.
  • Available on every Beatview plan, including Free. Plan limits still apply: publishing stops once you reach your active-job limit.
Who uses it

Built for agency desks. Open to any agent.

Recruiters on the desk

Triage new applicants, brief yourself before a call and draft client updates without switching tabs.

Agency owners and team leads

Ask across every open role: what is live, where it is stuck and who needs action today.

Teams building their own agents

An approved requisition opens the job and posts the link in Slack; a morning agent pulls the top-scoring new candidates for every role.
Build guide: agent + live dashboard
FAQ

Questions about the MCP server

Still deciding?

Bring a live vacancy and we’ll walk through where automation ends and recruiter review begins.

MCP (Model Context Protocol) is an open standard that lets AI assistants such as Claude and ChatGPT use tools and data from other software. The Beatview MCP server is the address your assistant connects to so it can work with the jobs and candidates in your Beatview account.

Any tool that supports remote MCP servers, including Claude, ChatGPT, Gemini and Cursor. Add https://platform.beatview.ai/mcp as a custom connector (Claude), custom app (ChatGPT and Gemini) or server entry (Cursor and other configuration-file tools), then approve access in Beatview. Each assistant has its own plan and region requirements, listed on its Beatview integration page.

Every plan, including the free plan. Plan limits still apply, so an assistant cannot publish more jobs than your plan’s active-job limit.

No. Beatview presents AI scores as signals to review, never as a hiring decision, and tells the assistant to confirm with you before it moves a candidate or sends an interview invitation. You decide who moves forward.

No. Recordings are never shared with a connected agent. The assistant receives Beatview’s assessment of each answer plus a link that opens the recording inside Beatview for signed-in members of your account. Beatview stores the AI’s assessment of each answer, not a word-for-word transcript.

Yes. Screening, stage rules and AI interview invitations run inside Beatview whether or not an assistant is connected. The assistant is a way to ask about that work and act on it, not a requirement for it.

If your assistant supports scheduled or recurring tasks, you can schedule a prompt such as “Every morning, list new candidates scoring above 80 across my open roles.” The schedule runs in your assistant; Beatview answers each request like any other.

The data an assistant asks for is returned to that assistant, so it is handled under your agreement with that AI provider as well as Beatview’s Privacy Policy. Connect only assistants your organisation has approved for candidate data, and review what each connection has done in the MCP Server settings.

No. Connecting is one address and one approval screen. Developers building their own agents can use the same server with an MCP client configuration file.

Set-up guides

Connect your assistant

Beatview + Claude

Add Beatview as a custom connector in Claude.

Explore

Beatview + ChatGPT

Publish Beatview as a custom app in your ChatGPT workspace.

Explore

Beatview + Gemini

Add Beatview as a custom Connected App in Gemini.

Explore

Recruiting MCP servers compared

Which recruiting tools work with AI assistants, and what each one exposes.

Explore
Keep exploring

What your assistant is connected to

Resume screening

Score every application against the criteria your consultant configures.

Explore

AI video interviews

Structured, timed interviews candidates complete on their own schedule.

Explore

Work Style assessment

Structured insight into how candidates prefer to work and collaborate.

Explore

Recruitment automation

Stage rules that move candidates and send invitations for you.

Explore
Try it on a live role

Open your next role from the chat.

Create a free Beatview account, connect your assistant and ask it to open a job. The free plan includes one active role.

Start for free
  • Included on every plan
  • Asks before it acts
  • Recruiters keep every decision

Page last reviewed by the Beatview team.