Search your candidate database with AI

Describe the person you need

Open AI Search below Candidates and describe the experience, skills, credentials, employers, or background you want to find. Press Enter or choose the round arrow at the end of the second line to see ranked matches from your organization without leaving the page.

The page names the organization being searched. Its centered starting view becomes a compact search bar above the results, so you can refine the same request while keeping the candidate list in view. While RecruitIQ reads resumes and ranks candidates, the loading message explains that a deeper search can take 30 seconds to a minute.

Add job and location context when it matters

Choose Add Job inside the main search field to open a searchable list of CRM jobs. Pick a role to combine its title, requirements, skills, work mode, and credentials with the search. You can search from the selected job alone or add optional written details to narrow it further.

The selected role stays visible in the search field and can be changed or removed from the same popover. For a job-only search, Job description, Notes, and Location chips replace the text-entry line, while the selected job starts the second line on the left and the round search arrow finishes that line on the right. The extra space between rows keeps those controls easy to scan. Text that was already entered stays visible instead of becoming a hidden search rule. Hover, focus, or choose any chip to open a compact preview that stays away from the screen edge. An empty chip says that source is not being used instead of implying hidden context.

RecruitIQ uses the complete saved job context: title, job description, requirements, internal notes, structured skills and credentials, and any optional recruiter request. It also distills those sources into candidate-relevant concepts so the search has both the original meaning and a normalized recruiting brief. Different wording is allowed; a candidate can reach review because their documented work means the same thing even when the resume does not repeat the job’s exact phrase.

The first pass reads structured resume evidence across the organization, including current and prior work descriptions, and builds a high-recall candidate pool. Exact-title overlap is not required. A deeper AI comparison interprets the job criteria and creates the final shortlist. Words such as “must,” “required,” or “minimum” in a job description do not silently remove a person: a short experience history, missing credential detail, or different license jurisdiction can remain visible as a Potential fit when the resume still contains relevant career evidence. Broad project-management or operations experience alone does not qualify someone for a role whose defining work is in another domain.

RecruitIQ resolves the saved job location, shows each available candidate distance from the job, and uses proximity as a modest ranking factor without letting it outweigh resume relevance. Open the Location chip to keep Any distance or choose a 10, 25, 50, or 100 mile hard radius from that job. Candidates without usable coordinates can remain in an Any distance search but cannot pass a selected radius. When you start writing a custom request, a separate Filter location chip appears so you can explicitly search from a different city, state, ZIP code, country, and distance.

Add phrases such as “without sales,” “exclude agency,” or “-temporary” when a profile must be left out. Explicit exclusions are applied before the remaining candidates are ranked.

Scan a job-specific candidate briefing

Each candidate now appears in a compact one-column row with two to four of the strongest resume-backed reasons that matter for the selected job. Every reason uses a short claim followed by concrete proof from a title, employer, responsibility, technology, experience total, or credential. If the resume contains a supported experience total, the reason states the concrete number of years, such as “7 years of project-management experience,” instead of saying only that the person meets or exceeds the request. A profile with thinner evidence receives fewer reasons instead of padded or flattering language.

A light, outlined row of readable skill chips appears after the reasons when the candidate has stored profile skills. RecruitIQ uses the explicit skills attached to the candidate first and falls back to structured resume-profile keywords when no explicit skill list exists. These chips summarize existing profile data; AI Search does not invent new skills for the card.

Results remain sorted by relevance and show one simple badge: green Good fit for direct, credible core-role evidence or yellow Potential fit for useful adjacent, incomplete, or unverified evidence. The primary card does not expose match percentages, internal criterion states, signal counts, or missing-keyword inventories. Recent roles and documented work take priority over standalone skills. A generic resume summary or isolated keyword cannot be the only reason a candidate appears.

The claim-and-proof bullets are static, selectively highlighted in yellow, and intentionally short so the card stays fast to scan; there is no extra click or hover panel on each reason. Every AI-written reason must cite one supplied resume source, and RecruitIQ rejects unknown sources or proof that is not grounded in that source. When a resume is available, the resume icon opens it in the existing RecruitIQ drawer without leaving the search results. When a job is selected, Add to job stays in the card action rail and adds only that candidate to that exact role after the recruiter chooses it.

Keep candidate data inside the signed-in organization

AI Search derives organization and role access from the signed-in RecruitIQ session. It searches only active candidates inside that organization, and a selected job is checked against the same organization before search begins. Literal recruiter exclusions and a selected distance radius are enforced outside the AI comparison. Job wording about experience, licenses, education, or credentials influences fit and ordering but does not become an invisible hard filter.

Each completed search also creates a bounded backend quality trace showing every job-context source received, the semantic criteria extracted from those sources, each candidate’s internal met, partial, unverified, or not-met assessment, the full-resume corpus and candidate-pool counts, the embedding and AI models and prompt versions used, include or exclude decisions, final fit and rank, and the resume source IDs supporting each returned reason. Those internal assessments help RecruitIQ improve search quality; recruiters see only the concise briefing. The trace is organization-scoped and does not store candidate email, phone, resume URLs, full resume documents, or embeddings.

New candidates and meaningful profile changes are also prepared for indexed search in the background. Work descriptions are now part of that searchable profile and content hash, so future indexed retrieval does not collapse a strong career fact into a shallow title or skill list.

Resume preview repeats the same organization check before RecruitIQ reads the file. Running a search alone does not edit candidates, assign them to a job, move a pipeline stage, or write the recruiting request into a candidate record. Add to job is a separate, explicit recruiter action and uses RecruitIQ’s existing job-assignment workflow. Results and mileage depend on the profile and location information currently available for each candidate.