77% of Open Roles Filled From Existing Applicants: How AI Talent Rediscovery Is Replacing High-Cost Agency Recruiting
By Chris Weinmann, Founder, OVI
Most enterprises are sitting on millions of qualified applicants they have already screened, interviewed, or received applications from — and doing nothing with them. Every time a new role opens, recruiters start from scratch: posting jobs, sourcing externally, paying agencies. The candidates who applied last quarter, last year, or for a different role entirely sit untouched in the CRM.
AI talent rediscovery changes that equation. Instead of processing only new inbound applications, these systems simultaneously screen thousands of historical applicants against every open requisition — surfacing qualified candidates the organization already knows but has forgotten.
Two enterprise case studies from Workday HiredScore illustrate how this works at scale, across different industries and hiring challenges.
AdventHealth: $68 Million Saved and 655 Nurses Recruited in Year One
AdventHealth operates 52 hospitals across nine US states, serving 7.8 million patients annually. When the system launched Workday HiredScore's AI Fetch in May 2023, it was facing a national nursing shortage projected at approximately 78,610 FTE nurses for 2025.
The results over the first 12 months (May 2023–May 2024) were substantial:
- $68 million reduction in agency nursing spend in the first year
- 655 qualified nurses recruited through AI Fetch leads
- 77% of nursing requisitions covered by existing past-applicant pools
- 100% increase in job requisitions closed — effectively doubling recruiter productivity
- 50% higher offer rate for AI-identified Fetch leads compared to other external candidates
- 40% decrease in hiring manager decision time from resume review to scheduling
The mechanism is straightforward. Rather than relying solely on new job-board applicants or expensive staffing agencies, AI Fetch scanned AdventHealth's historical applicant database — candidates who had previously applied, withdrawn, or been considered for different roles — and matched them against current open nursing positions. The AI surfaced an average of 300+ past applicants per open requisition who met the qualification criteria.
This did not replace recruiter judgment. Recruiters still reviewed, contacted, and selected candidates. What changed was the starting pool: instead of waiting for new applicants or paying agencies $15,000–$30,000 per travel nurse placement, recruiters had a pre-qualified pipeline of candidates who had already expressed interest in the organization.
AdventHealth's HR team described this as moving from a "post and pray" model to what they called an orchestrated talent acquisition approach, where AI continuously matches existing talent to emerging needs across the 52-hospital system.
Automotive Manufacturer: 80% Requisition Coverage at Global Scale
A major global automotive manufacturer — handling between one and two million candidates annually — deployed the same Workday HiredScore AI Fetch technology to address a different challenge: rapidly identifying candidates with EV and zero-emissions skill profiles from its historical talent database during a workforce transformation.
The outcomes paralleled AdventHealth's experience:
- 80% of open requisitions covered by existing talent pools
- 72% increase in screening efficiency
- 800% higher apply-conversion rate for AI-identified Fetch leads versus other external sourcing
- 220% higher offer rate for AI leads compared to external candidates
At this scale — millions of candidates processed annually — the productivity gains compound. Traditional keyword-based ATS searches cannot efficiently screen a database of one to two million historical applicants against hundreds of simultaneously open roles. AI Fetch performs this matching continuously, identifying candidates whose skills, experience, and qualifications align with new openings regardless of when or which role they originally applied for.
The automotive manufacturer's experience demonstrates that talent rediscovery is not industry-specific. Whether the challenge is filling nursing positions during a healthcare staffing crisis or sourcing EV engineers during an automotive transformation, the underlying problem is the same: qualified candidates already exist in the database but are invisible without AI-powered matching.
What Makes Talent Rediscovery Different From Standard AI Screening
The distinction matters for HR leaders evaluating AI hiring tools:
Traditional AI screening processes new inbound applications in real time. A candidate applies, the AI scores and ranks them, and the recruiter sees a prioritized shortlist. This is valuable but addresses only one side of the funnel — new applicants.
AI talent rediscovery simultaneously screens thousands of past applicants who withdrew, were not matched, or applied for a different role against every new opening. Most enterprises have hundreds of thousands — often millions — of historically qualified applicants sitting dormant in their CRMs and ATS platforms. Talent rediscovery makes this database actively searchable at scale.
The practical difference is enormous. AdventHealth found that 77% of its nursing requisitions could be filled from existing applicants. The automotive manufacturer hit 80%. These candidates did not need to be sourced, marketed to, or acquired through agencies — they were already in the system.
For organizations spending heavily on external sourcing and staffing agencies, talent rediscovery represents a direct cost reduction. AdventHealth's $68 million savings in the first year came almost entirely from reduced agency dependency.
Structured AI Evaluation: The Infrastructure That Makes Rediscovery Work
When AI talent rediscovery surfaces 300+ past applicants per open role, the evaluation layer becomes critical. Keyword matching alone cannot reliably assess whether a candidate who applied for a different position two years ago is qualified for today's opening.
OVI's Milo agent addresses this with rubric-weighted CV evaluation — configurable weights, context clues, and red flags — which is the structured evaluation layer that makes talent rediscovery work reliably at scale. When AI surfaces hundreds of past applicants per role, Milo-style rubric scoring ensures every candidate gets consistent, criteria-driven evaluation rather than keyword matching that misses transferable skills and relevant experience.
Frequently Asked Questions
What is AI talent rediscovery?
AI talent rediscovery is the practice of using artificial intelligence to screen historical applicant databases — candidates who previously applied, withdrew, or were considered for different roles — against current open positions. Unlike traditional ATS keyword searches, AI talent rediscovery systems can simultaneously evaluate thousands of past candidates against multiple open requisitions, identifying qualified matches that would otherwise remain dormant in the system.
How does AI talent rediscovery differ from a standard ATS?
A standard ATS processes and stores applications but typically only searches historical records through manual keyword queries. AI talent rediscovery actively and continuously matches past applicants to new openings using skills-based AI evaluation, not just keyword matching. This means a candidate who applied for a different role or at a different time can be automatically surfaced when a relevant position opens — without the recruiter needing to know they exist in the database.
What ROI can organizations expect from AI talent rediscovery?
Results vary by organization size, database depth, and industry, but the two enterprise case studies reported here show that 77–80% of open requisitions could be filled from existing talent pools. AdventHealth documented a $68 million reduction in agency spend in the first year. The key ROI drivers are reduced external sourcing costs, faster time-to-fill (candidates are already partially vetted), and higher offer acceptance rates for AI-identified candidates who have prior engagement with the organization.
What is AI talent rediscovery?
AI talent rediscovery is the practice of using artificial intelligence to screen historical applicant databases — candidates who previously applied, withdrew, or were considered for different roles — against current open positions. Unlike traditional ATS keyword searches, AI talent rediscovery systems can simultaneously evaluate thousands of past candidates against multiple open requisitions, identifying qualified matches that would otherwise remain dormant in the system.
How does AI talent rediscovery differ from a standard ATS?
A standard ATS processes and stores applications but typically only searches historical records through manual keyword queries. AI talent rediscovery actively and continuously matches past applicants to new openings using skills-based AI evaluation, not just keyword matching. This means a candidate who applied for a different role or at a different time can be automatically surfaced when a relevant position opens — without the recruiter needing to know they exist in the database.
What ROI can organizations expect from AI talent rediscovery?
Results vary by organization size, database depth, and industry, but the two enterprise case studies reported here show that 77–80% of open requisitions could be filled from existing talent pools. AdventHealth documented a $68 million reduction in agency spend in the first year. The key ROI drivers are reduced external sourcing costs, faster time-to-fill (candidates are already partially vetted), and higher offer acceptance rates for AI-identified candidates who have prior engagement with the organization.