Every closed requisition leaves behind candidates who almost got the offer. Six months later, a similar role opens. Most teams start sourcing from zero anyway.
That habit wastes budget on sourcing candidates who are already in your own pipeline history. Greenhouse offers recruiting teams tools to revisit past applicants. However, these tools do not include automatic, database-wide searching and matching.
In this guide, we'll explain what Greenhouse offers to find silver medalists and qualified candidates in your ATS database. It also covers where those native tools fall short, and how an AI integration fills the gap.
What Is Candidate Rediscovery?
Candidate rediscovery means resurfacing qualified candidates who are already in your talent pools and ATS database. It covers people from closed jobs, expired pipelines, and past sourcing efforts.
Your database becomes a real sourcing channel once past candidates get matched against new openings automatically. Most ATS platforms let you search history. Few score that history against a live job automatically.
Does Greenhouse Provide Candidate Rediscovery?
Yes, Greenhouse provides partial candidate rediscovery but only through manual filters, split across two separate tools. The Past Candidates tab, inside Talent Matching, searches for applicants who have already applied to a past job, filtered by scorecard rating and milestone reached. It does not search prospects, candidates sitting in a talent pool who were sourced but never applied.
For that, a recruiter has to switch to the older Talent Rediscovery tool and choose candidates or prospects one search at a time, not both together.
Neither tool assigns a match score against your current job's requirements. Both surface people using filters built from past hiring activity, not automated live scoring against this role. A recruiter also has to open the right tool and rebuild filters manually for every job, every time.
How to Rediscover Candidates in Greenhouse ATS
Greenhouse's native tools search on saved filters, not automated scoring against an open job. Most teams integrate Skima AI to search and score every matched candidate in the full database. Below is a 5-step workflow, starting with what Greenhouse already gives you.
Step 1: The Native Path, Start With Past Candidates
Open the job dashboard for your live requisition and click into Talent Matching. Select the Past Candidates tab. Greenhouse loads a suggested filter set automatically, built from signals like scorecard rating, milestone reached, and rejection reason, so there's already a starting shortlist instead of a blank table.
Four preset filter sets control how wide that shortlist is. Focused shows only 100% Yes scorecards from the last 12 months, Standard loosens to 75% Yes over two years, Broad drops to 50% Yes, and Broadest includes any rating. Widen or narrow based on how thin the results look.
That tab only surfaces people who already applied to a past job. It works on filters, not on a score, so nothing here tells you how well a candidate fits this job's actual requirements. That's the gap Skima AI's rediscovery closes next.
Step 2: Connect Your Greenhouse Account With Skima AI
Generate a Harvest API key, or set up OAuth, inside your Greenhouse account. This connection uses Greenhouse's own Harvest API directly, with no middleware required. Skima AI requires read access to candidates, applications, attachments, and jobs.
That access pulls context for scoring every candidate correctly. It also needs write access to Notes, custom fields, and Tags. That access sends results back inside Greenhouse once screening and scoring are complete.
- Read access: Application, Candidate, Attachment, and Job records
- Write access: Application custom fields, Candidate tags, and Activity notes
- Optional webhooks: Trigger scoring the instant an application updates or a stage changes
If webhooks are not enabled, Skima AI falls back to safe polling every 10 to 15 minutes.
Step 3: Configure Custom Fields and Confirm the Connection
Add the Application fields Skima AI writes to, or let Skima create them during setup: AI Match Score as a number from 0 to 100, Bullet Reasons as a short block of text, and a Shortlist Link. A SKIMA Screened tag gets added at the candidate level, so screened candidates are filterable in Application Review or anywhere else you already work.
Optional HM Feedback and HM Comment fields capture manager responses once you send a shortlist out. If your Greenhouse instance restricts custom fields, Skima AI falls back to writing an Activity Note or attaching a SKIMA_Evidence.pdf instead.
Before going further, test this on two or three requisitions: upload a resume, then confirm the Match Score and Reasons actually populate on that application.
Step 4: Launch Talent Rediscovery Against a Live Requisition
Open the job inside Skima AI, from the Jobs list or from within the job itself. Click the three-dot Actions menu and select Rediscover Candidates. A modal lets you scope the scan in two ways: limit it to a saved candidate segment, or leave it open to every candidate record Skima AI has synced from Greenhouse, including ones with no link to this requisition.
Set the Candidate Created Date range next: All time, Last 6 months, Last 1 year, or a custom window. Click Apply. Skima AI scans the selected pool against this job's specific requirements and sends a notification once the scan finishes. Refresh the page to see the results.
Step 5: Review, Filter, and Sync Feedback Back to Greenhouse
Open the View Candidates tab to see the ranked results, each with an AI match score and reason bullets (explaining the fit) tied to that score. Filter by Industry, Work Mode, Experience Level, Notice Period, Location, or Salary Range to narrow the list down.
Once you've built a shortlist, generate a secure, no-login link and send it to the hiring manager. They can click thumbs up, thumbs down, or maybe on each candidate, with an optional comment.
Skima AI writes that decision back into Greenhouse through the same API connection for the audit trail, updating the custom fields and logging an Activity Note, so the feedback remains on the record without anyone touching a second tool.
What If Your Greenhouse Database Is Messy, Duplicated, or Stale?
Greenhouse flags a possible duplicate whenever a new profile matches your tag settings. Those settings typically check for shared email, phone number, or LinkedIn URL. This tag appears on the candidate profile and inside Talent Matching results. Flagging a duplicate is not the same thing as fixing one, though.
Resolving a flagged duplicate still takes a manual merge from your team. Greenhouse warns this action cannot be undone once it is completed. Auto-merge handles tightly matched criteria, but broader matches still need manual review. Stale profiles compound the same problem as a database keeps growing. Candidates change numbers, update CVs elsewhere, and slowly go quiet.
Skima AI treats this as ongoing maintenance rather than a one-time project. It matches on email, phone, or LinkedIn URL first. It also checks matching name, location, and company together as a signal.
A side-by-side comparison appears before anything actually gets merged. Missing details get filled in automatically through contextual profile enrichment. A dedicated Duplicates tab logs every resolved case for later audit.
5 Benefits of Talent Rediscovery with Skima AI in Greenhouse
Rediscovery with Skima AI changes what a Greenhouse team can do with its own history, in several concrete ways.
- Faster Screening: Rediscovered candidates arrive already scored, so recruiters skip the raw resume pile entirely. In a recent Greenhouse pilot, a user reported time-to-screen reduced by 90%.
- More Interviews From the Same Shortlist: The same pilot saw far more interviews conducted from the top shortlist than before.
- Explainable Match Scores: Every rediscovered match carries an AI Match Score with reason bullets tied to skills and experience.
- Full Context Inside Greenhouse: That score and reasoning live directly on the Greenhouse record, so a recruiter reviewing a candidate months later skips reconstructing the logic.
- Instant Filtering: The Skima Screened tag makes these candidates instantly filterable within existing Greenhouse views.
- Hiring Manager Engagement: The shortlist feedback loop keeps hiring managers engaged without a separate login, and manager satisfaction increased significantly in that same pilot.
- Lower Cost Per Placement: For recruitment teams, this means placements from candidates already sourced once, reducing the cost of finding those same people again to almost nothing.
5 Best Practices to Rediscover Talent in Greenhouse
None of these results happen automatically inside Greenhouse without deliberate habits. The practices below cover data hygiene, scan timing, and fitting rediscovery around existing tools:
- Scan for Rediscovery Before Posting Anything New: Check Skima AI's results before writing a job ad or briefing a sourcing partner.
- Keep Your Duplicate Tag Criteria Tight: Review your Email, Phone, and LinkedIn settings regularly, since loose criteria let duplicates slip through.
- Segment Scans With the Candidate Created Date Filter: Choose a 6-month or 1-year window when a role needs someone recently active.
- For In-House Teams, Run Rediscovery Alongside Talent Matching: Use Talent Matching for live applicants and Skima AI for everyone else in the database.
- Turn on Automatic Rediscovery for Every New Job: Enable this setting so a scan starts the moment a new requisition opens.
Is Candidate Rediscovery Worth It for Your Recruitment Team?
The honest answer depends on how large and old your Greenhouse database is. A company with years of applications has real, unscored value sitting untouched.
Setup takes real effort too, and IT time matters here. A Harvest API key, custom field mapping, and webhooks all need attention upfront. Skima AI is also a paid layer sitting on top of Greenhouse. That cost remains on top of whatever tier already gates Talent Matching.
A newer company with a small, fast-moving pipeline has less to rediscover. For that team, the added cost may not pay off yet. Rediscovery earns its place once your database is large and aging.
Final Verdict: Pilot Candidate Rediscovery in Greenhouse Now
For most established Greenhouse customers, the case for piloting is strong. The right move is not a company-wide rollout on day one. Start with three live requisitions already struggling through normal sourcing.
Connect Skima AI, configure the custom fields, and launch rediscovery on those three roles. Structure the pilot around 30 days, with checkpoints at day 15 and day 30. That timeline matches what Skima AI's own Greenhouse integration recommends.
Track time-to-screen, how many rediscovered candidates reach an interview, and manager response. Based on results from similar pilots, expect faster candidate screening within that window. From there, expanding to more requisitions becomes a data-backed decision, not a guess.
Frequently Asked Questions
1. Does Greenhouse have a built-in dedicated candidate rediscovery feature?
Partially, Greenhouse splits this across two tools: Past Candidates, which searches applicants only, and the older Talent Rediscovery, which searches candidates or prospects separately. Neither tool scores a person against your current job's requirements.
2. How to conduct candidate rediscovery in Greenhouse ATS?How to conduct candidate rediscovery in Greenhouse ATS?
Connect Skima AI to Greenhouse through a Harvest API key or OAuth. Then open any job and launch Talent Rediscovery, which scans every synced candidate against that job's actual requirements and returns a ranked, filterable shortlist.
3. What are the benefits of Greenhouse candidate rediscovery?
In a documented 30-day Greenhouse pilot, Skima AI cut time-to-screen by 90% and doubled interviews from the shortlist. Hiring manager satisfaction rose by 24 points on NPS, driven by the no-login shortlist feedback loop.
4. Is Skima AI safe to integrate with Greenhouse for talent rediscovery?
Yes, Skima AI encrypts candidate data in transit and at rest, operates under a signed Data Processing Agreement, and maintains SOC 2 compliance. Match scores rely on resume-based evidence only, with no automatic rejections at any stage.
5. Is candidate rediscovery useful for Greenhouse recruiting teams?
Yes, especially for the corporate in-house teams that make up most of Greenhouse's customer base. The value scales with database size and age, so a company with years of applicant history gains more than a newer one.
