Get Started

Quickstart: Connect Your AI Assistant to SeekOut

Connect any MCP-compatible AI assistant to SeekOut in minutes and start searching talent, reviewing your pipeline, and taking action from chat.

If you have a SeekOut Recruit license, MCP access is already included — there's nothing extra to buy. On a self-service assistant like Claude, you can be running your first search in about five minutes.

Prerequisites

A SeekOut Recruit license (MCP access is included — no upgrade needed)
An MCP-compatible AI assistant: Claude (all tiers), ChatGPT, Gemini Enterprise, Microsoft 365 Copilot, Amazon Quick Enterprise, or a desktop app like Claude Desktop, Cursor, or VS Code

Enterprise-managed platforms (Gemini Enterprise, Microsoft Copilot) may require an administrator to provision the connector before individual users can connect.

Connect in Four Steps

Choose your AI assistant

Pick any supported assistant. You can connect more than one — each uses your same SeekOut account and permissions.

Add the SeekOut MCP connector

Use the MCP Server URL when adding the connector:

https://seekout-search-mcp.seekout.io/tools

For desktop apps, add this to your MCP config file instead:

{
  "mcpServers": {
    "seekout-search": {
      "url": "https://seekout-search-mcp.seekout.io/tools"
    }
  }
}

Adding the connector differs slightly by assistant — follow the short setup guide for your tool: Claude, ChatGPT, Gemini Enterprise, Microsoft Copilot, Amazon Quick, or desktop apps.

The first time your assistant calls a SeekOut tool, you'll be prompted to sign in. OAuth handles authentication automatically — API keys are only needed for automated, non-interactive environments.

Run your first search

Ask in plain English. Any of these is a great first prompt:

Find senior product managers in Seattle with fintech experience.

How many ML engineers are available in Austin vs. Denver vs. Chicago?

Pick Your Assistant

Each assistant has a short, dedicated setup guide:

Claude

Add SeekOut as a custom connector on Free, Pro, Max, Team, or Enterprise.

ChatGPT

Install the published SeekOut app — no Developer Mode needed.

Gemini Enterprise

SeekOut-provisioned setup with your Google Cloud admin.

Microsoft Copilot

Add SeekOut as a federated connector in the M365 Admin Center.

Amazon Quick

Auto-registers via Dynamic Client Registration — no manual keys.

Desktop Apps

Claude Desktop, Cursor, VS Code, and other config-file clients.

Five Scenarios for Your Week

Once you're connected, try these. Each maps to a guided workflow — copy the prompt, change the variables, and use the follow-up to refine.

Analysis & Intelligence01 / 05

Analyze Talent Markets

Use case · Hiring Manager Reset

Give me a talent market analysis for Staff ML Engineer in Boston with 8+ years. Show pool size, top employers, common skills, and which requirements are narrowing the field. Tell me what would expand the pool if we need to broaden the spec.

Then try · Compare Boston with Seattle and Austin at the relaxed spec. Where does the supply widen most?

Talent market analysis dashboard rendered live in Claude

Real result in Claude — market dashboard: pool size, seniority mix, locations, top titles, employers & skills. Click to enlarge.

Weeks of stalled req → market truth in one chatBest for TA Leaders, Hiring Managers

Search & Discovery02 / 05

Source Across Pools

Use case · Cross-Pool Shortlist Build

Find senior React engineers in NYC with 5+ years and TypeScript expertise — one ranked list with match reasons. If our ATS is connected, fold in anyone already in our pipeline too.

Then try · Show me the top 10 with a one-line summary of what makes each strong.

Ranked candidate shortlist rendered live in Claude

Real result in Claude — ranked candidate shortlist with match reasons and one-click actions. Click to enlarge.

Minutes, not MondayBest for Sourcers, Full-cycle Recruiters

Outreach & Messaging03 / 05

Reach Out

Use case · Personalized Candidate Outreach

Draft warm outreach to the top React candidate. Reference one real detail from their background, keep it senior-IC in tone, and give me a couple of angles to choose from — show me the draft before any send.

Then try · Tighten it to three sentences, and draft a short LinkedIn InMail version too.

Personalized outreach email drafted in Claude with two angles to choose from

Real result in Claude — a personalized outreach email (two angles), grounded in the candidate’s profile and ready to open in your mail client. Click to enlarge.

From shortlist to sentBest for Sourcers, Full-cycle Recruiters

Evaluation & Hiring04 / 05

Assess Candidates

Use case · Candidate Fit Assessment

Score the top candidate against our Senior Platform Engineer role on (1) platform depth in production (AWS, Kubernetes, IaC), (2) senior-level scope, (3) market fit. Compare side-by-side with three alternates.

Then try · Re-run at the 8–15-year band and prioritise anyone with explicit search-infrastructure or vector-DB experience. Compare to today's top 4.

Real result in Claude — candidate profile scored against the role for a side-by-side comparison. Click to enlarge.

POV in minutesBest for Full-cycle Recruiters, TA Leaders

Action & Export05 / 05

Activate Shortlists

Use case · Outreach-Ready Workflow

Take action on the top 3 from our React shortlist. Verify contacts and save them to a new workspace called Senior React — NYC Q2 2026 so they’re ready for outreach.

Then try · Export the saved shortlist to a sheet I can share with the hiring manager.

React shortlist in Claude with verified contacts and one-click save to a workspace

Real result in Claude — act right on the results: verified contacts and a one-click save to the “Senior React — NYC Q2 2026” workspace. Click to enlarge.

Loop closed in one chatBest for Full-cycle Recruiters, Sourcers

Next Steps

Workflows

The 20 built-in workflows your assistant runs automatically.

Candidate Actions

Contact, outreach, export, ATS, and workspace actions.

Prompting Patterns

Six habits that get better answers from your assistant.

← Introduction Claude →