Sign in

Candidate Sourcing

Scheduled multi-platform candidate sourcing across LinkedIn, GitHub, Indeed, Dice, and niche boards, delivering ranked, tier-coded pipelines to Google Sheets with summary emails.

  • AI-assisted
  • Talent Acquisition
See it in action
Reduction in sourcing labor per requisition
89%
Increase in recruiter sourcing capacity
2-3x
Increase in sourcing surface area per role
3x

Role spec in. Sheet is ready.

See the sourcing agent take a Staff Machine Learning Engineer role from schedule to a ranked sheet.

Run CS-0417

Weekly refresh · Staff Machine Learning Engineer

  • AI
  • 89% less sourcing labor
  1. Intake
  2. Organize
  3. Evaluate
  4. Deliver

Stage 1 · Intake

Weekly run scheduled

Role spec uploaded, runs every Monday at 7am

New case

Run scheduled
Requisition
Staff Machine Learning Engineer · Open 6 weeks
Owner
Head of Talent · Summary email recipient
Schedule
Weekly, Monday 7am
Request
Ranked passive candidate pipeline · Low inbound applicant volume
Case packet · 3 sourcesComplete

Incoming

  • Role specSkills, years, level, locationReceived
  • ScheduleEvery Monday at 7amReceived
  • RecipientHead of Talent emailReceived

Stage 2 · Organize

Profiles merged

Four platforms searched, duplicates merged

Pulled together

Pool built
  • LinkedIn and GitHub searchedGitHub contributors to relevant open-source ML frameworks
  • Indeed and Dice sweptActive candidates added to the same pool
  • Duplicates mergedMatched by normalized name and current company

Matched record

Match key
Normalized name + current company
Found on
LinkedIn, GitHub
Flag
Multi-Source Verified
Profile URLs
Kept for each source

Stage 3 · Evaluate

Candidates tiered

Five scoring dimensions, Tier 1/2/3 labels

Checks applied

Scored 0 to 10
  • Skills matchRequired skills weighted high, preferred mediumRole spec
  • Experience and seniorityYears of experience and seniority alignmentRole spec
  • Location and title fitLocation fit and role title proximityScoring weights
Deciding finding

Tier set by fit score

A Fit Score of 8.0 or higher places a candidate in Tier 1 for immediate outreach.

Tier definitions

Tier 1
8.0 and up · Immediate outreach
Tier 2
6.0 to 7.9 · Review within 5 days
Tier 3
4.0 to 5.9 · Future roles

Stage 4 · Deliver

Pipeline delivered

Google Sheet written and summary email sent

Run complete

Ranked sheet and summary sent

Every candidate is scored on five dimensions and tiered, so the Head of Talent starts Monday with a shortlist to share.

Google Sheet
Scores, tiers, source URLs, notes
Summary email
Tier breakdown, top 5 Tier 1
Next run
Monday 7am

Delivered to

Head of Talent

Ready for the hiring manager

Replaces

4-6 hours

Manual sourcing per week

Audit trail

  1. Scheduled run startedMonday 7am07:00
  2. Four platforms searchedLinkedIn, GitHub, Indeed, Dice07:04
  3. Candidates scored and tiered07:09
  4. Sheet written, summary sentHead of Talent07:11

89% less sourcing labor per requisition

One requisition, end to end. Illustrative case, sample data.

Use cases.What starts it, what it does, what you get.

  1. Use case 1

    Weekly Pipeline Refresh for a Hard-to-Fill Engineering Role

    Starts when

    A Staff Machine Learning Engineer role has been open for 6 weeks with low inbound applicant volume. The Head of Talent schedules weekly Monday runs.

    1. Head of Talent uploads the role spec and configures the weekly Monday 7am schedule with their email as the recipient.
    2. On the first run, the agent searches LinkedIn, queries GitHub for contributors to relevant open-source ML frameworks, and sweeps Indeed and Dice.
    3. The agent deduplicates all results by normalized name and current company, merges profiles found on multiple platforms, and flags them as 'Multi-Source Verified'.
    4. Each candidate is scored across five dimensions (skills match, years of experience, seniority alignment, location fit, role title proximity) and assigned a Tier 1/2/3 label.
    5. The agent creates a new Google Sheet, writes all candidates with scores, tier labels, source URLs, and notes, and applies conditional formatting.
    6. A summary email is sent listing total candidates found, tier breakdown, top 5 Tier 1 profiles, and source performance.

    Every Monday the Head of Talent receives a fresh, ranked shortlist of passive candidates with a formatted Google Sheet ready to share with the hiring manager, replacing 4-6 hours of manual sourcing per week.

  2. Use case 2

    Urgent Healthcare Staffing Surge, ICU Nurse Pipeline in 24 Hours

    Starts when

    A sudden staffing gap requires the recruiter to build a qualified ICU Registered Nurse pipeline for two Chicago-area facilities within 24 hours.

    1. The recruiter sends the agent a sourcing request with the role spec: ICU Registered Nurse, Chicago metro area, 3+ years ICU experience required.
    2. The agent prioritizes NurseRecruiter and Hirebridge as niche healthcare sources, then supplements with LinkedIn and Indeed.
    3. For each source, the agent logs the exact query used. Any source returning zero results is recorded with the query.
    4. Candidates are scored with location fit carrying standard 10% weight (non-remote role), with skills match weighted toward ICU-specific experience.
    5. The agent deduplicates across all four sources, flags multi-source verified candidates, and writes the pipeline to a Google Sheet.
    6. Because this is urgent, the agent checks whether at least 5 candidates scored above 4.0. If not, it appends a 'Criteria Broadening Recommended' section.

    Within hours, the recruiter receives a Google Sheet with a tiered ICU nurse pipeline and a summary email with actionable suggestions for broadening the search if needed.

Works with your systems

The systems this agent reads from and writes to, connected inside your environment.

  • Slack (sourcing run summaries)
  • ATS MCP Server (Greenhouse / Lever / iCIMS)
  • RequestApproval (HITL gate for sensitive roles)

See it in action.A conversation that moves the work forward.

Sample scenarios built in ACTAVA. Actions and outputs are illustrative.

Agents › Candidate Sourcing

Illustrative example

What does this agent do and how do I kick off a sourcing run?

Candidate Sourcing

I source candidates on a schedule across several platforms. To start a run I need a role spec, and I will create a new Google Sheet and send an email when the run completes.

What happens at completion

  • A new Google Sheet is created for the run
  • A summary email is sent when the run completes

Example complete

1 of 4

Where this agent works.The industries it serves, and customers doing similar work.

Put this agent to work on your cases.

We'll show it running against your own workflow, connected to your systems, with your team approving what matters.