AI Resume Screening for Ashby: What's Built In and When to Add Weighted Ranking

How Ashby's AI-Assisted Application Review Works

Ashby is an all-in-one recruiting platform: jobs, pipeline stages, scorecards, interview scheduling, and analytics. Its native AI screening feature, AI-Assisted Application Review, is documented in Ashby's knowledge base and is available on all Ashby plans. Here is how it works, based on Ashby's public docs:

1. Opt in. An Org Admin enables AI features (accepting Ashby's AI terms) and then turns on AI-Assisted Application Review under Admin > Organization Setup > Opt-In Features.
2. Define criteria per job. In the job's AI Features settings you add resume criteria: suggestions generated from the job posting, criteria reused from other jobs, or custom criteria with a short title and a prompt. You can also check whether long-form application questions were answered (that check confirms an answer exists, not its quality).
3. Validate. Ashby checks each criterion for clarity and objectivity and scans for implicit or explicit Equal Employment Opportunity (EEO) issues, flagging anything that needs a rewrite.
4. Evaluate. “Save and Evaluate all active candidates” runs the AI over the pipeline. Each criterion comes back as Meets, Does Not Meet, or Undecided, with reasoning and citations on the candidate's Feedback tab.
5. Review and sort. In the Application Review stage you can add an Evaluations column and an AI Job Criteria Met Percentage column, sort by it, filter by specific criteria with the AI Criteria Quick Filter, and run bulk actions on the filtered list.

Ashby also gives reviewers ways to stay in control: you can flag an evaluation as incorrect or misleading, re-evaluate all active candidates after editing criteria, and configure an automated-processing legal notice that lets candidates opt out of AI evaluation. Ashby's docs note that its human-reviewer Average Score is calculated separately from AI criteria evaluations.

Pricing is credit-based: per Ashby's documentation, one AI credit equals one evaluation of a single candidate, supporting up to 50 criteria, and each plan includes a credit allowance with more available in blocks. Check Ashby's current docs for your plan.

Source: Ashby Knowledge Base, “AI-Assisted Application Review” and “Enabling AI features in Ashby” (docs.ashbyhq.com), checked October 2026. Ashby features change; confirm details with Ashby.

How to evaluate Ashby's AI-assisted candidate screening and application review

Test it on one live role. Write five to eight singular, resume-verifiable criteria, validate them, and evaluate all active candidates. Sort by AI Job Criteria Met Percentage, read the reasoning on Undecided results, and check whether the top ten match the hiring manager's picks. If one criterion matters far more, compare a weighted ranking on the same pack.

1. Pick one role with real volume and a hiring manager who will review the top 10.
2. Write criteria the way Ashby recommends: singular, specific, objectively verifiable. Fix anything flagged in validation before you evaluate.
3. Evaluate, add the AI Job Criteria Met Percentage column, sort.
4. Spot-check: Undecided results, a few Does Not Meet results near the top, and the citations.
5. Measure time to shortlist, hiring-manager acceptance of the top 10, and whether known-good profiles rose.
6. If one or two criteria clearly matter more than the rest, rank the same pack in SkipCV with weights and deal-breakers and compare the two top sets.

Custom scoring weights per competency, without coding

Ashby's public docs describe a Meets / Does Not Meet / Undecided result per criterion and a criteria-met percentage. They don't describe per-criterion weights, and Ashby says its AI does not rank or give numerical ratings. In SkipCV you set how much each competency matters, per role, in the app. You need no code, no integration and no admin ticket, and true must-haves stay separate as deal-breakers. See weighted criteria matching.

Fast shortlisting and bulk actions for engineering hiring managers

In Ashby you can filter Application Review with the AI Criteria Quick Filter and run bulk actions on the filtered list. That suits hard yes/no checks. When an engineering role has one critical stack skill and several nice-to-haves, a weighted ranking puts the critical skill first, and every rank comes with a reason a hiring manager can check. See bulk resume screening.

Where Criteria-Met Checks Stop

Ashby's approach is transparent and well suited to clear, verifiable checks. Ashby's own guidance is to write criteria that are singular, specific, and objectively verifiable from a resume. The trade-off is that each criterion is a yes/no judgement, and the sortable column is described as the percentage of criteria a candidate met.

That works when every criterion matters about the same. It gets harder when they do not. If a role has one critical skill and five nice-to-haves, a candidate who meets the five nice-to-haves can sort level with, or above, someone who has the critical skill. You can work around that with filters on specific criteria, but you are then building the weighting by hand, role by role.

Weights express how much something matters; deal-breakers express whether an applicant can be considered at all. The full approach is in candidate screening software with weighted criteria matching, and the general buyer criteria are in our guide to applicant screening software: what it is and what to look for.

When Ashby's AI Review Is Enough, and When to Add a Layer

Keep Ashby — Add a Screening Layer Where It Helps

Most Ashby buyers shopping for an “alternative” do not actually want to migrate jobs, scorecards, interview kits, and analytics. They want better shortlist quality. That is the fork covered in our SkipCV vs Ashby comparison and the broader ATS alternative for recruiters guide: replace workflow software only when the ATS itself is broken; otherwise extend it.

SkipCV does not use a native Ashby connector; it works alongside the ATS. If you want to see the scoring loop step by step, read how resume screening software works.

High-Volume Applications Are the Real Test

Any screening approach looks fine on ten resumes. The test is a popular role that floods overnight. If your stack cannot run bulk resume screening at pack scale, with resume #200 judged on the same criteria as resume #1, it will not fix the bottleneck — it will only rename it.

Buying Checklist for Ashby Teams

For a wider market map, see our candidate screening software comparison and AI resume screening software guide. Teams already on Greenhouse or Lever may prefer the parallel playbooks for AI resume screening for Greenhouse or AI resume screening for Lever.

Why SkipCV Fits Ashby Shops

SkipCV is built as a recruiter-first screening layer for recruiters and TA teams: bulk packs, weighted scoring, deal-breaker filters, transferable-skills recognition, and ranked shortlists with transparent fit reasons — without replacing Ashby as your hiring OS for jobs, pipeline, and analytics.

The easiest way to judge it is on your own applicants: create a quick free account and try it — no obligations whatsoever. New accounts include 20 free credits, enough to rank a live Ashby role and compare the result with Ashby's criteria-met view. See SkipCV pricing when you get there.

FAQ

Read These Next