Using MCP Server with AI Interviewer & Notetaker

With the Manatal MCP Server, you can connect your AI assistant (such as Claude or ChatGPT) to your Manatal account and work directly with your AI Interviewer assessments and AI Notetaker transcripts and meeting summaries. Instead of opening each candidate one by one, you can ask your assistant to compare candidates, find the evidence behind a rating, spot and fill in information missing from a profile, or build a report for your client, all in plain language.

This section explains what you can access and walks through the most useful ways to put it to work.

Before you start

• Connect the Manatal MCP Server to your AI assistant. Learn more in this guide.
• AI Interviewer: Assessments and transcripts are available once a candidate has completed their AI interview and the assessment has been generated.
• AI Notetaker: Transcripts and meeting summaries are available for meetings the Notetaker joined, usually within a minute after the meeting ends.
  Note

  Your calendar must be integrated with Manatal for the Notetaker to join meetings.

• Permissions: Your assistant can only access the candidates, jobs and interviews you can see in Manatal. Access follows your Manatal user permissions.

What you can access

Because the MCP Server also gives access to candidate and job pipeline data, your assistant can combine interview content with candidate profiles and the jobs they are applying for.

AI Interviewer Use Cases

1. Compare candidates against one another

When several candidates have completed the AI interview for the same job, ask your assistant to put them side by side. It reads each assessment and transcript and highlights where candidates are stronger or weaker on the criteria that matter for the role.

Example prompts

• Compare the candidates who completed the AI interview for the Senior Accountant job. Rank them by overall rating and summarize the main strengths and concerns for each.

• For the Sales Manager job, which candidates scored highest on negotiation skills? Show me a table with their ratings for each grading criterion.

• Between Sarah Lee and James Tan, who is the better fit for the Customer Success Lead role, and why?

Tips

• Name the job so your assistant compares candidates on the same interview and grading criteria.
• Ask for a table if you want a quick, scannable comparison.

2. Get deeper evidence behind a rating

A rating tells you how a candidate performed. The transcript tells you why. Ask your assistant to pull out the exact parts of the conversation that support (or contradict) a given rating, so you can validate the assessment before moving a candidate forward.

Example prompts

• Why did Maria Lopez get a 4/5 on communication? Quote the parts of her AI interview transcript that support this rating.

• Find the moments in John Smith's AI interview where he talks about leading a team. Do they support his leadership rating?

• Is there anything in this candidate's transcript that contradicts their high rating on technical skills?

Tips

• Ask for direct quotes from the transcript so you can see the candidate's own words.
• Use this before a hiring manager review, or when a rating looks surprisingly high or low.

3. Create a custom report to share with your client

Turn AI Interviewer results into a client-ready document in the format you want. Your assistant can summarize each shortlisted candidate, include their ratings and key quotes, and structure the report the way your client prefers.

Example prompts

• Create a shortlist report for my client on the Marketing Director job. Include the top 3 candidates by AI interview rating, a short summary of each, their ratings per criterion, and one supporting quote per candidate.

• Write a one-page summary of Anna Chen's AI interview that I can send to the hiring manager. Keep it professional and focus on her experience with B2B SaaS.

• Build a comparison table of all candidates who completed the AI interview for this job, with a recommendation at the end.

Tips

• Tell your assistant who the report is for and what format you need (email, one-pager, table, slides).
• Review the report before sending it. Check that it only includes information you are happy to share with your client.

AI Notetaker Use Cases

1. Detect and add information missing from the candidate's profile

Candidates often share useful details during interviews that never make it into their profile: salary expectations, notice period, availability, location preferences, skills or past experience. Ask your assistant to read the Notetaker transcripts and flag what's missing or out of date compared with the candidate's profile. Once you've reviewed the findings, your assistant can update the candidate profile in Manatal for you.

Example prompts

• Read the transcripts of my meetings with David Kim and list any information he mentioned that isn't in his candidate profile.

• Did this candidate mention their notice period or salary expectations in any of our calls? If so, what did they say?

• Compare what Priya said in her interview with what's on her resume. Is there anything new or inconsistent?

• Add David's notice period and salary expectations from our last call to his candidate profile.

Tips

• Ask your assistant to list each finding with the quote it comes from, so you can verify before updating the profile.
• Use the meeting summary for a quick overview, and the full transcript when you need exact details.
• Check the proposed changes before asking your assistant to update the profile, so only verified information is saved.

2. Compare candidates across their completed interviews

When several candidates have been interviewed in meetings the Notetaker joined, ask your assistant to compare them based on what was actually said. This is especially useful when interviews were run by different recruiters or hiring managers.

Example prompts

• Compare the three candidates I interviewed this week for the Product Designer job. Summarize how each answered questions about their design process.

• Across all interviews the Notetaker joined for the Operations Manager job, which candidates showed the most relevant experience with supply chain tools?

• Create a table comparing these candidates on experience, motivation, availability and salary expectations, based on their interview transcripts.

Tips

• Make sure the Notetaker joined each interview you want to compare. Meetings without the Notetaker won't have a transcript.
• Give your assistant the criteria you care about so the comparison is consistent across candidates.

Combining AI Interviewer and Notetaker

If a candidate completed an AI interview and then had a follow-up call which the Notetaker joined, you can ask your assistant to look at both together.

• Summarize everything we know about Tom Nguyen from his AI interview and our follow-up call. What questions are still open before we make a decision?

Tips for better results

• Be specific: name the candidate, the job, or the timeframe you're interested in.
• Ask for sources: request quotes from the transcript to see exactly where an insight comes from.
• Review before you share: AI assistants can make mistakes. Check key facts against the assessment or transcript in Manatal before sending anything to a client or making a hiring decision.

FAQ

Q1. Which AI assistants can I use?
ChatGPT and Claude. Refer to the MCP Server guide for setup instructions.

Q2. Can my assistant access interviews that are still in progress?
No. AI Interviewer data is available once the interview is completed and assessed. Notetaker transcripts and meeting summaries are available once the meeting has ended, usually in under a minute.

Q3. Does my assistant see all candidates in my account?
No. Access follows your Manatal user permissions, so your assistant only sees the same data you can access in Manatal.