PitchMe adds AI assistant search and shortlist tools for recruiters

PitchMe launched a remote MCP server that lets recruiters search, shortlist and reveal candidate data from ChatGPT, Groq and Muse, then push results into Bullhorn or Airtable. The move is designed to cut steps between sourcing and placement while keeping agency permissions and candidate data controls in PitchMe.

Why it matters: - PitchMe is trying to move recruiters from prompt-based research to in-workflow action. - The new MCP server connects AI assistants to verified candidate data, which can speed up sourcing and shorten the path to outreach and placement. - For agencies, the system is meant to keep searches, shortlists and permissions inside the agency’s PitchMe account.

What happened: - PitchMe launched a remote Model Context Protocol server that works with ChatGPT, Groq and Muse. - Recruiters can describe a role in plain language, ask an assistant to search PitchMe, build a shortlist and surface available contact details. - The launch is live now for PitchMe customers. - Agencies can request a demonstration at pitchme.co. - Setup guides are available at the MCP server documentation.

The details: - A recruiter can ask for a role-specific search, such as finding ICU nurses in Texas with active state licenses, shortlisting the top 20 and revealing their contacts. - The assistant runs the search in PitchMe and returns a shortlist the recruiter can use. - The assistant works only within the recruiter’s account, team and permissions in PitchMe. - Users can set the connection for a limited time or keep access until they revoke it. - Users can disconnect the assistant at any time. - PitchMe is SOC 2 certified and GDPR compliant. - For healthcare staffing, the assistant can search the US national provider dataset, NPPES. - Matched healthcare records can show NPI number, credentials, specialty and license states. - The update also adds Shortlist & Reveal, which lets recruiters save an external candidate from search results and request contact details in one step. - Recruiters can run reveals across a whole project shortlist. - Bullhorn and Airtable write-back lets recruiters send revealed candidates to an open Bullhorn job or Airtable table after integration and field mapping are configured. - Candidates already on that job or table are excluded automatically. - The release also adds CSV and Excel candidate imports. - PitchMe added self-service team joining through verified company domains. - Search coverage expanded in Belgium, Bulgaria, Luxembourg and the Netherlands.

Between the lines: - PitchMe is positioning AI as an execution layer, not just a drafting tool. - The company is also leaning on governance features, including account-level permissions and revocable access, to address agency control concerns. - Healthcare staffing appears to be a priority use case, given the NPPES integration and matching fields. - The release extends PitchMe’s broader SONAR strategy, which ties an agency’s internal database to the external talent market.

What's next: - Agencies using ChatGPT, Groq or Muse can now test the workflow inside PitchMe. - The company says setup requirements vary by assistant, so implementation will depend on the AI platform and the agency’s existing integrations. - PitchMe will likely push adoption through demos, documentation and its existing ATS connections. - The launch follows SONAR’s debut at Staffing Industry Analysts’ CollaborationX conference in Dallas in September.

The bottom line: - PitchMe is betting recruiters want AI that can do the sourcing work, not just describe it, while keeping candidate data and downstream ATS updates inside agency controls.