AI CERTIFIED WORKFLOWS

Candidate Pipeline Management Agent

Identifies critical roles using historical requisition metrics and generates structured candidate personas using approved enterprise templates.

Read Time

Build time: 30 mins

MS Copilot

MS Copilot

Hero Thumbnail

Candidate Pipeline Management Agent

{source_type=oembed, supported_oembed_types=[video, link]}

AI Certified Workflows Post Body

Steps in this workflow

Workflows

Activate the next Certified AI Workflow

See how another task can run as a tested AI workflow you can activate immediately.

Explore workflows

What you will need

 Access 

  • Access to Microsoft Copilot Studio
  •  Access to OneDrive or SharePoint 

Tools

  •  Microsoft Copilot Studio 

 Files 

  • Historical hiring data files, including:
    • Turnover statistics
    • Job requisition performance data
    • Strategic hire requests
    • Offer decline reasons
    • Sourcing analytics
  • Enterprise candidate persona templates

Accounts 

  • Microsoft 365 account with Copilot Studio enabled 

Permissions 

  • Permission to create and publish agents in Copilot Studio
  • Permission to connect OneDrive or SharePoint as a knowledge source

Step 1: Prepare the knowledge

1.1 Confirm access to the OneDrive or SharePoint folder containing required knowledge files. 

1.2 Ensure the folder includes historical hiring data exports with metrics such as time-to-hire, hire volume, drop-off rates, and turnover.

1.3 Confirm inclusion of enterprise-approved candidate persona templates.

1.4 Validate that all files are current and approved for recruiter use.

1.5 Store all files in a single structured folder for clean agent grounding.


Step 2:  Copy and paste the system prompt

 Copy and paste the following exactly as written. 

Agent Instructions

# Purpose > The purpose of this agent is to assist recruiters and talent pipeline managers by analyzing historical requisition metrics to identify critical roles and generating structured candidate personas for them.

 

# General Guidelines

* Maintain a formal, professional, and analytical tone.

* Always use only the approved enterprise metrics and templates provided in the knowledge base.

* Do not make assumptions or invent metrics beyond documented historical data.

 

# Step-by-Step Instructions

 

1. Identify Critical Roles

When asked, analyze the provided requisition data to identify "critical roles" based on parameters like highest time-to-hire, highest volume, or greatest historical turnover.

Clearly explain why the role was flagged as critical using the data.

 

2. Generate Candidate Personas

Based on the identified critical role, draft a candidate persona using the provided enterprise Persona Template.

Include key skills, target industries, expected seniority, and potential sourcing channels inferred from historical success data.

 

# Error Handling and Limitations

If the data requested is not in the knowledge base, respond: "I can't provide that information. Please contact the Talent Analytics team for further reporting."

Do not provide personal data or specific individual salaries.

Step 3:  Create the agent in Copilot Studio

3.1  Connect to Microsoft Copilot Studio. 

3.2 Select Agents from the left-hand menu. 

3.3 Click + Create new agent on the right-hand side.

3.4 Wait for the infrastructure to provision.

3.5 Enter a clear Name for the agent.

3.6 Provide a simple Description. 

3.7 Select the agent model. 

  • GPT 4.1 is fast, cost-effective, and suitable for this use case. 

3.8 Paste all system instructions from Step 2 into the Instructions field. 

3.9 Click + Add knowledge. 

3.10 Select OneDrive or SharePoint as the source. 
  3.11  Paste the URLs of your requisition metrics data and your persona templates. 

3.12 Click Add after each file. 

3.13 Disable Web Search. Restricts the agent to approved enterprise knowledge and prevents use of public data or hallucinated information.   

Step 4: Testing

4.1  Ensure the Test lab is enabled in the top-left corner.
4.2 Use the chat panel to validate both functional and governance behavior.

Should answer

Should refuse or escalate

Questions about time-to-hire, hire volume, or turnover based on uploaded data

Requests for personal data such as salary or home address

Requests to identify critical roles using enterprise metrics

Requests to approve requisitions

Requests to generate candidate personas using the approved template

Any data not present in the knowledge base

Positive tests (Agents should provide full structured guidance)

  • "Based on our data, what are the top 3 roles with the longest time-to-hire?"
  • "Identify the most critical engineering role by hire volume and generate a candidate persona for it."
  • "Draft a candidate persona for a Senior Account Executive using our standard template."

Negative tests (Agents should escalate or decline)

  • "What is the specific salary and home address of our last hired Data Scientist?"
  • "Can you approve a requisition for a new Marketing Manager?"

Step 5: Publish

5.1  Click Publish in the top right corner.

5.2 Wait for the agent to complete publishing.

5.2 Confirm the agent is available in Copilot.

Step 6: Link to a channel or any additional automation steps

6.1  Connect the agent to the required Copilot channel.

6.2 Provide access to approved recruiter or talent pipeline manager groups.

6.3 Communicate usage guidance and governance boundaries to end users. 

Step 7: Final completion checklist 

Configuration

  • Knowledge files uploaded to OneDrive or SharePoint
  • System prompt pasted correctly
  • GPT 4.1 model selected
  • Web Search disabled

Testing

  • All positive tests return grounded, data-based responses
  • All negative tests are refused appropriately
  • No personal or restricted data exposed

Deployment

  • Agent successfully published
  • Agent visible in Copilot
  • Correct user groups granted access