Eightfold AI and External Sourcing: More Candidates for the Matching Engine to Work With

Eightfold's AI matching and talent intelligence tools are built to find the best-fit candidates from whatever data the platform has access to — internal employees, past applicants, and any external candidate pools connected to it. That last part matters: the matching engine can only surface candidates that exist in its data. It doesn't independently generate new passive candidates from outside the platform.

AI matching still needs a candidate supply

For roles where the internal and historical applicant data is thin — new skill sets, new markets, highly specialized or highly passive candidate segments — Eightfold's matching quality is limited by how few relevant candidates it has to compare against. More relevant data in generally means better matches out.

Connecting an external sourcing layer

Eightfold supports API-based candidate ingestion, allowing structured profiles to be added to the platform's talent pool. A sourcing layer that runs targeted social campaigns to reach passive candidates in specific skill areas can pre-screen and deliver them into Eightfold, expanding the pool the matching engine works from.

What it means for a talent intelligence strategy

• Matching quality improves for skill sets and markets where internal data was previously thin
• New, currently-passive candidates enter the system instead of relying only on past applicant history
• Recruiters get more relevant AI-surfaced matches instead of the same limited internal pool
• Existing matching models, workflows, and internal mobility processes remain unchanged

When this is worth setting up

This matters most for skill sets or markets that are new to the organization, where there's little internal or historical applicant data for the matching engine to draw on yet. Established roles with years of internal data benefit less from external sourcing than from the existing model.

What SmartHire adds above Eightfold

SmartHire sources and pre-screens candidates through targeted social campaigns, then feeds qualified profiles into Eightfold via API — expanding the data the matching engine has to work with. Your existing models, workflows, and internal mobility strategy stay exactly as they are.

Checklist before connecting to Eightfold AI

• Which jobs or regions in Eightfold AI have the thinnest candidate flow today?
• Which fields should come with each candidate, such as contact details, screening answers and source?
• Who owns access, permissions and the integration user?
• How will duplicates be detected, and how long are records kept?
• Which numbers will you compare before and after: cost per qualified candidate, time to first candidate, recruiter hours?

A simple way to pilot it

Start with one role in one location, ideally roles where the existing talent data has too few matches. Run the sourcing layer for a few weeks alongside your normal process, and compare it with the same role's inbound results. If the numbers hold, extend to more roles. This is a pilot, so the results depend on the role and market, and you should treat any published benchmark as a guide only.

Candidate data is personal data. Check that consent, data-processing terms and storage locations meet your rules before you go live, and confirm what your plan and version of the system allow.

Does Eightfold AI need to be replaced or changed?

No. The sourcing layer runs alongside it, and your team keeps working in Eightfold AI exactly as before. It is a good fit for organisations that use AI matching on their talent data.

How long does setup take?

It depends on which fields you connect and what your IT team needs to approve. Starting with a single role keeps the first setup short and makes problems easy to spot.

How do we know it is working?

Compare cost per qualified candidate, time to first qualified candidate and recruiter hours on screening against your existing baseline for the same type of role.