> ## Documentation Index
> Fetch the complete documentation index at: https://docs.talview.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Gen AI Resume Fitment: Feature Overview

***

### What is Gen AI Resume Fitment?

Gen AI Resume Fitment is Talview's AI-powered screening layer that automatically evaluates incoming candidate resumes against the specific competency and skill requirements of a workflow (job requisition) - and returns a structured fitment verdict with an evidence-backed explanation.

It is designed to remove subjectivity from early-stage screening, surface the right candidates faster, and give recruiters a consistent, auditable signal - at scale.

***

### How It Works

1. **Configure** - Recruiter or Master Recruiter sets Essential, Preferred, and Optional criteria against the workflow as per the job role requirements
2. **Parse** - System extracts structured data from each incoming resume across four dimensions:
   | Dimension                           | What is Extracted                                | Used For                                                                                  |
   | :---------------------------------- | :----------------------------------------------- | :---------------------------------------------------------------------------------------- |
   | **Personal Information**            | Name, contact details, location                  | Candidate identification and profile creation; not used in fitment scoring                |
   | **Education Details**               | Degrees, institutions, years of completion       | Evaluated against Essential criteria (e.g., minimum degree requirement)                   |
   | **Experience Details**              | Job titles, companies, tenures, responsibilities | Primary signal for Essential and Preferred criteria - seniority, recency, domain depth    |
   | **Qualifications & Certifications** | Listed certifications, professional credentials  | Evaluated against Essential or Preferred criteria where specific credentials are required |
3. **Evaluate** - LLM-backed engine returns Strong / Potential / Low Fit with evidence-backed explanation, visible on the candidate card. Please check the ***demo*** below:

<div style={{ position:"relative",paddingBottom:"calc(54.02777777777777% + 41px)",height:0,width:"600px",margin:"0 auto" }}>
  <iframe src="https://demo.arcade.software/wAAc3Xewpk5WxobzpJyc?embed&embed_mobile=tab&embed_desktop=inline&show_copy_link=true" title="Resume Shortlisting Demo" frameBorder="0" loading="lazy" webkitAllowFullScreen mozAllowFullScreen allowFullScreen allow="clipboard-write" style={{ position:"absolute",top:0,left:0,width:"100%",height:"100%",colorScheme:"light" }} />
</div>

***

### Configuration Criteria

| Criteria Type                | Definition                                                              |
| :--------------------------- | :---------------------------------------------------------------------- |
| **Essential** (Must-have)    | Non-negotiable requirements the candidate must demonstrate              |
| **Preferred** (Good-to-have) | Attributes that strengthen candidacy; not disqualifying if absent       |
| **Optional** (Nice-to-have)  | Supplementary signals that add positive weight; absence never penalises |

***Example - Senior Backend Engineer (5+ years)***

| Categories    | Criteria                                                                                      |
| :------------ | :-------------------------------------------------------------------------------------------- |
| **Essential** | Java or Python, 5+ years backend, distributed systems, B.E./B.Tech, Tier-1 graduation college |
| **Preferred** | AWS/GCP, AI-led development, microservices, team leadership                                   |
| **Optional**  | Open-source contributions, fintech domain, product company background                         |

***

### Fitment Tier Definitions

| Tier              | When It Applies                                                                                                                 |
| :---------------- | :------------------------------------------------------------------------------------------------------------------------------ |
| **Strong Fit**    | All essentials fully met with recent, unambiguous evidence; most preferred met. <br />Ready to contribute with minimal ramp-up. |
| **Potential Fit** | All essentials met (at most one partially); at least 40% of preferred met. <br />Minor gaps addressable through interview.      |
| **Low Fit**       | One or more essentials not met, or two+ essentials only partially met. <br />Material domain, seniority, or skill mismatch.     |

***Example: Output - Candidate Level***

For a candidate evaluated against the *Senior Backend Engineer (5+ years)* workflow above, the system returns:

**Fitment:** Potential Fit

**Explanation:**

*Strengths:* Strong backend engineering foundation across Java and distributed systems with progression from associate to senior engineer roles.

*Gaps:* Explicit backend tenure of three years versus five-year requirement; no cloud platform or leadership evidence in resume.

*Verdict Rationale:* Essential experience requirement only partially met; warrants interview to validate cloud exposure and depth of backend ownership.

***

### What Makes Talview’s Resume Fitment Distinctive

| Capability                | Generic AI Screeners   | Talview Gen AI Resume Fitment                           |
| :------------------------ | :--------------------- | :------------------------------------------------------ |
| Customer-defined criteria | Pre-trained taxonomies | Per-workflow Essential / Preferred / Optional           |
| Explainability            | Score or summary       | Evidence-traced verdict per criterion                   |
| Hallucination guardrails  | Variable               | Evidence-only; flags gaps over inflated scores          |
| Bias controls             | Often opaque           | No protected attributes used; criteria-based evaluation |
| Compliance posture        | Variable               | Aligned to EU AI Act high-risk system requirements      |

***

### Trust & Governance

* **Evidence-grounded** - No inferences beyond explicit resume content; if a criterion cannot be verified, the system states the gap rather than assuming
* **Bias-controlled** - Evaluates only against your configured criteria; protected attributes (gender, race, age, ethnicity, disability) are not used
* **Compliance-aligned** - Designed for EU AI Act high-risk system standards: transparent reasoning, configurable criteria, mandatory human oversight, and auditability
* **Human-in-the-loop** - The AI classifies and explains; recruiters and hiring managers retain full override authority

***

### Frequently asked questions

<Accordion title="Does the AI automatically reject candidates?">
  No. The AI classifies and explains; all shortlisting decisions are made by the hiring team.
</Accordion>

<Accordion title="Can I upload resumes in bulk? ">
  Yes. Resumes can be uploaded individually or in bulk. Candidate profiles are auto created, and fitment runs automatically once criteria are configured.
</Accordion>

<Accordion title="Can criteria be changed mid-cycle?">
  Yes. Criteria can be updated at any time and re-applied to existing candidates in the workflow.
</Accordion>

<Accordion title="Is candidate data used to train the model? ">
  No. Talview does not use individual customer data for model training without explicit consent and appropriate anonymisation.
</Accordion>

***

*For configuration support or fitment calibration queries, contact your Talview Customer Success Manager.*