BrightHire MCP

Overview

The BrightHire MCP (Model Context Protocol) Connector allows AI tools to securely access and work with your BrightHire data. By connecting an MCP-compatible client (like Claude, Cursor, or others), you can query interview data, generate summaries, and analyze hiring signals directly within your AI workflows.

The Model Context Protocol (MCP) is an open standard that enables AI tools to connect to external data sources in a structured and secure way.

With the BrightHire MCP connector, AI tools can retrieve and reason over:

BrightHire can be connected directly within Claude, making setup quick and configuration-free.

Once connected, Claude can access your BrightHire data to answer questions, generate summaries, and support hiring decisions.

Most MCP-compatible clients support OAuth-based authentication. This is the simplest way to connect, as users authenticate directly with their BrightHire account—no API key required.

Add the BrightHire MCP endpoint to your MCP client configuration:

Each MCP client handles configuration differently. Some require editing a local config file, while others provide MCP settings in the UI. Refer to your client’s documentation for exact instructions.

When your MCP client connects for the first time, it will initiate an OAuth flow. A browser window will open prompting you to log in to your BrightHire account and approve access.

After approval, the client receives an access token that allows it to retrieve data on your behalf.

Once authentication is complete, your MCP client can begin requesting structured data from BrightHire.

You can now:

The BrightHire MCP connector enables deeper analysis and workflows across your interview data using AI tools. Common use cases include:

These workflows allow teams to move faster and make more informed hiring decisions directly within their AI tools.

The MCP connector provides access to data stored in BrightHire, including:

Access through MCP follows the same permission model as BrightHire. Users can only access data they already have permission to view.

🔍 Search & Discovery

search — The main entry point. Search interviews using filters like candidate name, interviewer, position, department, stage, date range, outcome, call type, and more. Returns interview metadata including call IDs you can pass to other tools.

get_scheduled_interviews — Fetch upcoming scheduled interviews (defaults to next 7 days). Can filter by candidate name or position title, and scope to just your own interviews or your whole org.

📋 Interview Prep

get_interview_context — Get full prep context for a scheduled interview: start time, candidate/position metadata, interview guide, and candidate summary. Takes a scheduled_interview_id.

get_interview_guide — Retrieve the interview guide(s) for one or more scheduled interviews (up to 20 at once).

📝 Post-Interview Content

get_interview_summary — Get a post-interview summary for a single interview: interviewer notes, emoji reactions (thumbs up/down, stars, flags), and action links (BrightHire URL, share URL, ATS scorecard URL).

get_call_ai_notes — Fetch AI-generated Q&A pairs extracted from the transcript for up to 20 interviews at once. Much lighter than full transcripts. Supports filtering by keyword or speaker type (candidate vs. interviewer).

get_transcripts — Pull full verbatim transcripts and user-created call notes for up to 20 interviews. Large payload — best used when you need exact wording rather than summaries.

get_candidate_summary — Get the most recent candidate summary notes for a candidate.

📊 Evaluation & Scoring

get_scorecards — Retrieve ATS scorecards (overall recommendation, attributes, questions) for up to 20 interviews. Can query by call_ids or by candidate_id + position_id.

get_job_description — Get the job description for a position using its position_id (found in search results). Useful for evaluating candidate fit against actual role requirements.

🏋️ Coaching

get_coaching_report — Returns your personal AI coaching report as markdown, summarizing trends and feedback from your recent interviews. No arguments needed — it infers you from the session.

If you have any problems connecting your MCP client to BrightHire, please reach out to us.

Hiring decisions often involve multiple stakeholders (recruiting, hiring managers, execs), each with different priorities. The BrightHire MCP connector can help align everyone by synthesizing interview data into a shared, consistent narrative.

Instead of manually stitching together feedback, AI tools can pull insights across interviews and present a unified view of candidate strengths, risks, and tradeoffs tailored to different audiences.

What you can build:

A stakeholder-ready briefing that automatically adapts the same candidate data for different audiences (e.g., exec summary for leadership, detailed evaluation for hiring panels), ensuring everyone is aligned before final decisions.

Beyond evaluating candidates, interview data contains signals about candidate intent, motivation, and potential offer risk. By analyzing language across interviews, you can identify candidates who may be hesitant, have competing priorities, or need a more tailored close.

This shifts MCP from just evaluation → conversion optimization.

What you can build:

A pre-offer risk report that highlights signals like hesitation, competing offers, compensation expectations, or misalignment—along with recommended talking points for recruiters to improve close rates.

Your interviews contain a huge amount of institutional knowledge—about roles, technologies, competitors, and candidate expectations. MCP allows you to systematically capture and reuse that knowledge.

This turns interviews into a continuously growing knowledge base.

What you can build:

A searchable knowledge layer across all interviews, allowing teams to quickly answer questions like:

Example: How are BrightHire Engineering Candidates talking about AI?

Review all interview data for candidate Joe Brown and generate an offer risk and close strategy brief:

Include: