# Yemez: full text

> The same text as the pages on yemez.com, generated 2026-09-30. Vuplicity’s published package prices, read from its public catalog on September 29, 2026: Basic $7.99, Essential $14.99 and Complete $24.99. Package prices are not an all-in quote. Court and data-access fees can vary by jurisdiction.

---

# Yemez: background checks your AI agents can order

URL: https://yemez.com/

Building AI hiring software?

# Background checks your AI agents can order.

1.  [Read the catalog.](https://yemez.com/connect#read)
2.  [Order in sandbox.](https://yemez.com/connect#order)
3.  [Get the status back.](https://yemez.com/connect#events)

Add the background check step to your hiring product. The candidate gives consent in a hosted flow on their own phone, and a consumer reporting agency prepares the check. Today, that is Vuplicity.

[Explore the integration](https://yemez.com/connect) [Talk to Utlyze](https://www.utlyze.com/?book=call) [Paste to your AI agent](https://yemez.com/connect#paste)

Yemez is the platform. Vuplicity is the consumer reporting agency that prepares checks today. Hiring decisions stay with the employer.

One background check, start to finish

Illustrative

AI agent stops here AI agent stops here

1.  Your AI agent Laptop Reads the catalog, places the order for the employer and reports the status back into your product.
2.  Vuplicity, the CRA CRA cabinet The consumer reporting agency prepares and delivers the background check. Today, Vuplicity.
3.  The candidate Consent phone The candidate reviews the disclosure and gives consent in the CRA’s hosted flow, on their own phone. AI agent stops here.
4.  The employer Report folder The report goes from Vuplicity to the employer. Hiring decisions stay with the employer. AI agent stops here.

Your AI agent

Reads the catalog, places the order for the employer and reports the status back into your product.

Follow the handoffScroll or use the slider 

Who takes each step

1.  Ordered by **Your AI agent** For an employer with a permissible purpose, such as employment screening.
2.  AI agent stops here: the candidate’s sensitive details
    
    **Sensitive details never go into the chat.** The CRA sends the candidate a secure link, and the candidate enters them there.
    
    Illustrative chat inside your hiring product
    
    1.  Hiring manager
        
        Their date of birth is Masked
        
    2.  Your product’s AI agent
        
        Please don’t share that here. Casey enters it in Vuplicity’s secure link.
        
    
    Vuplicity publishes this rule for AI agents that order its checks. [See all five rules](https://yemez.com/connect#rules)
    
3.  Consent **The candidate** (a person acts here) In the CRA’s hosted flow, on their own phone.
4.  Prepared by **A consumer reporting agency** Today, Vuplicity.
5.  Status **Back in your product** At your offer and onboarding stages.
6.  AI agent stops here: the hiring decision
    
    **The AI agent never makes or implies a hiring decision.** It reports status. The employer reads the report and decides. If they consider adverse action, the [pre-adverse and adverse action steps](https://www.ftc.gov/business-guidance/resources/background-checks-what-employers-need-know) apply.
    
    Vuplicity publishes this rule for AI agents that order its checks. [See all five rules](https://yemez.com/connect#rules)
    
7.  Decision **Your customer, the employer** (a person acts here) Hiring decisions stay with the employer.

Basic $7.99 Essential $14.99 Complete $24.99

Package price, not all-in. Court and data-access fees can vary by jurisdiction.

Read from Vuplicity’s public catalog on Sep 29, 2026, 19:53 UTC Fulfilled by Vuplicity

Illustration: one background check from order to decision, with the two places your AI agent stops.

-   MCP server: live, read-only
-   Orders: REST API with an organization key
-   Consent: hosted link on the candidate’s phone
-   Sandbox first, live after approval
-   Fulfilled by Vuplicity

## Your AI agents already source, screen and schedule candidates. Now they can order the background check too.

1.  Done: Source candidates
2.  Done: Screen applicants
3.  Done: Schedule interviews
4.  New: Order the background check
    
    -   Goes to a consumer reporting agency. Today, Vuplicity.
    -   Consent link goes to the candidate’s phone.
    -   Report goes to the employer. The employer decides.
    

Illustration: the hiring tasks your AI agents already handle, and the new one.

## Who does what

Yemez is the platform. It does not prepare or furnish consumer reports. A consumer reporting agency does that, and the two decisions that matter stay with people.

Illustration Five roles. Two human decisions. Select an object to inspect its role.

1.  ### Your AI agent
    
    Reads the catalog, places the order for the employer and reports the status back into your product.
    
    Who does what
    
    ### Your AI agent
    
    Reads the catalog, places the order for the employer and reports the status back into your product.
    
    Close
    
2.  ### Yemez
    
    Shows your AI agents how to order from a CRA and points them to the CRA’s live connection details. Utlyze runs it and takes every call.
    
    Who does what
    
    ### Yemez
    
    Shows your AI agents how to order from a CRA and points them to the CRA’s live connection details. Utlyze runs it and takes every call.
    
    Close
    
3.  ### The CRA
    
    Prepares and delivers the background check. Today, that is Vuplicity.
    
    Who does what
    
    ### The CRA
    
    Prepares and delivers the background check. Today, that is Vuplicity.
    
    Close
    
4.  AI agent stops here
    
    ### The candidate
    
    Decides on consent
    
    Reviews the disclosure and gives consent through a secure link on their phone.
    
    Who does what
    
    ### The candidate
    
    Decides on consent
    
    Reviews the disclosure and gives consent through a secure link on their phone.
    
    Close
    
5.  ### The employer
    
    Decides on hiring
    
    Reads the report and makes the hiring decision.
    
    Who does what
    
    ### The employer
    
    Decides on hiring
    
    Reads the report and makes the hiring decision.
    
    Close
    

## Consent happens on the candidate’s phone, not in the chat.

Your AI agent sends the candidate a secure link from the CRA. The candidate reviews the disclosure, gives consent and enters their sensitive details there, not into the chat and not to Yemez. Tell them who asked for the check and which CRA prepares it, so they know the request is real.

Illustration A handoff to a person

Explore the screens on the phone.

Your hiring product AI agent’s workspace

   ![](/art/consent-status/laptop.webp)

Illustrative chat inside your hiring product

1.  Hiring manager
    
    Casey accepted our conditional offer. Please start their background check.
    
2.  Your product’s AI agent
    
    Starting an Essential check for employment screening. I’ll send Casey a secure link from Vuplicity to review the disclosure and give consent.
    
3.  Hiring manager
    
    Their date of birth is Masked
    
4.  Your product’s AI agent
    
    Please don’t share that here. Casey enters it in Vuplicity’s secure link.
    

The chat stays here. Scroll to read earlier messages.

AI agent  
stops here

The candidate takes over.

The candidate’s phone CRA-hosted flow

Preview candidate screens. Illustration only.  Disclosure  Authorization  Details

   ![](/art/consent-status/phone.webp)

Example Co. Hiring

Disclosure

Your background check for the Operations Analyst offer

Example Co. asked Vuplicity, a consumer reporting agency, to prepare a background check about you. You’ll review the disclosure and give consent in Vuplicity’s secure link.

Open the secure link

Authorization

Consent is the candidate’s choice.

The candidate reviews the disclosure and gives consent in Vuplicity’s hosted flow.

A person gives consent.

The AI agent waits here.

Details

Entered on the phone.

The candidate enters their sensitive details in the CRA’s hosted flow.

Example candidate

Casey

Date of birth

Masked

These details stay out of the chat.

Requested by:

Example Co.

Prepared by:

Vuplicity, a consumer reporting agency

Questions about the check:

contact Vuplicity

Questions about the job:

contact Example Co.

Illustration only

Sensitive details stay here.

Illustrative. The notice comes from your product. The disclosure and consent screens are hosted by the CRA.

[How consent works, step by step](https://yemez.com/how-it-works#consent)

## Get the status back, inside your product.

Show every state where your team already works. Vuplicity sends lifecycle events to your server, and your product maps them to its offer and onboarding stages.

Illustration A status ticket in your product

Explore a state. Each stop shows who acts next.

1.  Waiting on consent The candidate (a person acts here)
    
    `PENDING_CONSENT`
    
    **Next:** The candidate reviews and consents.
    
    Waits on a person
    
2.  In progress The CRA
    
    `PENDING` `PROCESSING`
    
    **Next:** The CRA runs the check.
    
    Runs
    
3.  More information needed The candidate (a person acts here)
    
    `NEED_INFO`
    
    **Next:** The candidate adds information.
    
    Waits on a person
    
4.  Report ready The employer (a person acts here)
    
    `COMPLETE`
    
    **Next:** The employer reviews in the portal.
    
    Waits on a person
    
5.  Needs attention Your team
    
    `FAILED`
    
    **Next:** Your team resolves the exception.
    
    Runs
    

Status comes back to your product. The employer reads the report in Vuplicity’s portal.

Illustrative states, including waiting and exception paths.

Illustrative. Status names come from Vuplicity’s developer docs; the state labels are examples for your product. Vuplicity’s sandbox plays clear, record found, need info and dispute scenarios, so you can build each screen before a live check.

[See the full status map](https://yemez.com/how-it-works#status)

## Five rules your AI agent follows

Vuplicity publishes these rules for AI agents that order its checks. Two of them mark where your AI agent stops.

Illustration The rules, pinned where the work happens.

1.  **Order a check only when the employer has a permissible purpose, such as employment screening.**
    
2.  **Don’t collect the candidate’s sensitive details in chat.** Send the hosted consent link instead. It handles the disclosure and authorization.
    
    AI agent stops here
    
3.  **Don’t make or imply a hiring decision.** Report results to the employer. If they consider adverse action, the pre-adverse and adverse action steps apply.
    
    AI agent stops here
    
4.  **Quote prices from the live catalog, not from memory.** Package prices aren’t all-in. Court and data-access fees can vary by jurisdiction.
    
5.  **Treat guidance as guidance, not legal advice.** Vuplicity is not legal counsel.
    

[Rules and endpoints on Connect](https://yemez.com/connect#rules)

## Priced by the CRA, read from its live catalog

Each check is priced by the consumer reporting agency that prepares it. These are Vuplicity’s published package prices, read from its public catalog when this site was built.

Illustration Published catalog

Feed receipt Reset receipt

The receipt is fully fed. The package prices and searches are in the adjacent table.

Priced by the CRA that runs the check.

Vuplicity Catalog snapshot

Vuplicity’s published package prices
| Package | What it searches | Published price |
| --- | --- | --- |
| Basic | National criminal database, Sex offender registry, Global watchlist, Address history | $7.99 |
| Essential | All Basic checks, County and state court searches (upfront) | $14.99 |
| Complete | All Essential checks, Federal criminal search | $24.99 |

Package prices are not an all-in quote. Court and data-access fees can vary by jurisdiction.

Read from Vuplicity’s public catalog on September 29, 2026.

In the package price

The searches listed above.

Not in the package price

Court and data-access fees. Vuplicity finalizes them once the check type and the jurisdictions in the candidate’s seven-year address history are known.

Read from Vuplicity’s public catalog on Sep 29, 2026, 19:53 UTC. Fulfilled by Vuplicity [See Vuplicity’s pricing](https://www.vuplicity.com/pricing)

## Where to go next

-   ### Building an ATS or HRIS?
    
    Open door Close door
    
    Map each check’s status to your offer and onboarding stages, and keep the employer’s permissions as they are.
    
    [See the status map](https://yemez.com/how-it-works#status)
    
-   ### Running staffing software or a worker marketplace?
    
    Open door Close door
    
    Settle who the employer is, who pays and who reads the report before your AI agents order.
    
    [See the four questions](https://yemez.com/how-it-works#who-is-the-employer)
    
-   ### Run a consumer reporting agency?
    
    Open door Close door
    
    Talk with us about AI agents ordering from your agency.
    
    [For CRAs](https://yemez.com/for-cras)
    
-   ### Hiring for your own business?
    
    Open door Close door
    
    Ask before you build anything. Tell Utlyze how you hire, and we’ll help you work out the next step.
    
    [Talk to Utlyze](https://www.utlyze.com/?book=call)
    

## Add the background check to your hiring product.

Tell us how your product hires today. Utlyze runs Yemez and takes every call.

[Talk to Utlyze](https://www.utlyze.com/?book=call) [Or explore the integration](https://yemez.com/connect)

---

# How an AI agent orders a background check | Yemez

URL: https://yemez.com/how-it-works

1.  [Home](https://yemez.com/)
2.  [How it works](https://yemez.com/how-it-works)

# The life of one background check

Who does what, from the employer’s request to the hiring decision. Two places are marked where your AI agent stops and a person takes over.

[Explore the integration](https://yemez.com/connect) [Talk to Utlyze](https://www.utlyze.com/?book=call)

The human handoff Illustration

   ![A candidate’s phone on a warm white hiring desk. A physical marigold ticket says AI agent stops here at the consent handoff.](/3d/lifecycle/desk-04.webp)

**A person takes over.** Consent stays on the candidate’s phone. The hiring decision stays with the employer.

The seven steps

## Follow one check across the desk.

One continuous hiring desk Illustration

     ![A physical hiring desk with an employer folder, AI agent laptop, candidate phone, court searches, report release tray and employer decision folder. Marigold stop tickets mark both human handoffs.](/3d/lifecycle/desk-00.webp)

01 / 07 **The employer asks for a check.**

An employer starts the check with a permissible purpose.

[Read this step](https://yemez.com/how-it-works#lifecycle-step-1) Scroll or choose a step

Illustrative sequence, fulfilled by Vuplicity. No live check.

1.  The employer (a person acts here)
    
    ### Step 1: The employer asks for a check.
    
    A check starts with an employer that has a permissible purpose, such as employment screening. The employer also follows the state and local laws that apply to background checks and hiring, including laws that limit when and how they may consider criminal history.
    
    [The FTC’s guidance for employers](https://www.ftc.gov/business-guidance/resources/background-checks-what-employers-need-know)
    
2.  Your AI agent
    
    ### Step 2: Your AI agent places the order.
    
    It reads the live catalog first, so it quotes current prices. Then your server sends the order to Vuplicity’s REST API with the employer’s organization API key. In sandbox, nothing reaches live data providers or live billing.
    
3.  AI agent stops here: the candidate’s sensitive details
    
    **Sensitive details never go into the chat.** The CRA sends the candidate a secure link, and the candidate enters them there.
    
    Vuplicity publishes this rule for AI agents that order its checks. [See all five rules](https://yemez.com/connect#rules)
    
4.  The candidate (a person acts here)
    
    ### Step 3: The candidate gives consent.
    
    The CRA sends a secure link. The candidate reviews the disclosure and gives consent on their own phone. Their sensitive details go into that hosted flow, never into your AI agent’s chat. In sandbox, the email is suppressed, so you open the returned link yourself. Show this waiting state in your product.
    
    `PENDING_CONSENT`
    
5.  The CRA
    
    ### Step 4: The CRA checks the order.
    
    Vuplicity checks the order against the package and the rules for where the candidate lives and works.
    
    `PENDING`
    
6.  The CRA
    
    ### Step 5: The CRA searches the sources.
    
    Depending on the package: county and state courts, federal courts, a national criminal database, sex offender registries and global watchlists. If the CRA needs more from the candidate, the check waits in the hosted flow until they add it.
    
    `PROCESSING` `NEED_INFO`
    
7.  The CRA
    
    ### Step 6: The report is ready.
    
    The employer reads the report in Vuplicity’s portal. Your product gets the status through a webhook event.
    
    `COMPLETE`
    
8.  AI agent stops here: the hiring decision
    
    **The AI agent never makes or implies a hiring decision.** It reports status. The employer reads the report and decides. If they consider adverse action, the [pre-adverse and adverse action steps](https://www.ftc.gov/business-guidance/resources/background-checks-what-employers-need-know) apply.
    
    Vuplicity publishes this rule for AI agents that order its checks. [See all five rules](https://yemez.com/connect#rules)
    
9.  The employer (a person acts here)
    
    ### Step 7: The employer decides.
    
    Hiring decisions stay with the employer. If they consider adverse action based on the report, the [pre-adverse and adverse action steps](https://www.ftc.gov/business-guidance/resources/background-checks-what-employers-need-know) apply. A candidate who disagrees with the report can [contact the CRA that prepared it (Vuplicity’s summary of FCRA rights)](https://www.vuplicity.com/fcra-rights) to see their file or dispute information in it.
    

The location step, up close

## Where the check runs matters.

Vuplicity checks the order against the package and the rules for where the candidate lives and works.

   ![A three-dimensional US relief map on an off-white desk, with schematic Alaska and Hawaii insets and example location sockets in California, Colorado and New York.](/3d/jurisdiction/map-base.webp)

Illustration

Example locations. Simplified map with Alaska and Hawaii insets.

Choose an example location to raise its pin.

**California** Example home address Where the candidate lives is part of Vuplicity’s order check. **Colorado** Example address history The candidate’s address history helps establish the jurisdictions for the check. **New York** Example work location Where the candidate works also matters when Vuplicity checks the applicable rules.

Court searches depend on the package and locations. Hiring rules can also vary by state, county and city.

[Read the guidance below](https://yemez.com/how-it-works#guidance-title)

## Waiting and exceptions

A check waits on people and on sources. Design these screens before your first live check.

What each state means, and who acts next
| State | What it means | Who acts next | What your product can show (example) |
| --- | --- | --- | --- |
| Waiting on consent | The link is sent. The candidate has not finished consent. | The candidate | “Waiting on Casey” |
| More information needed | The CRA needs more from the candidate, in the hosted flow. | The candidate | “Casey needs to add information” |
| Records found | The report lists records for the employer to review. | The employer | “Report ready for review”, the same label as a clear report |
| Dispute | The candidate disputes information in the report with the CRA. | The CRA | A state that tells the hiring team the CRA is reviewing it |
| Needs attention | The check stopped and needs exception handling. | Your team | “Needs attention” |

Illustrative labels. Vuplicity’s sandbox has scenarios for clear, record\_found, need\_info and dispute. Check the event details for each in Vuplicity’s sandbox docs.

[Vuplicity’s sandbox docs](https://www.vuplicity.com/developers#sandbox) Fulfilled by Vuplicity

## Map the status into your product

Vuplicity sends lifecycle events to your server. Map each status to a stage your team already uses, and keep the employer’s permissions as they are.

Vuplicity statuses mapped to example stages
| Vuplicity status | What happened | Example stage in an ATS | Example stage in a worker marketplace |
| --- | --- | --- | --- |
| `PENDING_CONSENT` | The link is sent; waiting on the candidate | Offer accepted: waiting on background check consent | Onboarding: waiting on the worker |
| `PENDING` | Consent is in; the CRA checks the order | Background check in progress | Screening in progress |
| `PROCESSING` | The CRA is searching the sources | Background check in progress | Screening in progress |
| `NEED_INFO` | The CRA needs more from the candidate | Waiting on the candidate | Waiting on the worker |
| `COMPLETE` | The report is ready for the employer | Report ready for review | Report ready for review |
| `FAILED` | The check stopped | Needs attention | Needs attention |

Status names from Vuplicity’s developer docs. Stage names are examples.

[Webhook events on Connect](https://yemez.com/connect#events)

## Staffing or a marketplace? Settle who the employer is first.

When a staffing firm or a worker marketplace sits between the worker and the business, answer four questions with your customer before your AI agents order anything.

1.  Who is the employer asking for the check, with the permissible purpose?
    
2.  Whose account with the CRA places the order?
    
3.  Who reads the report?
    
4.  Who sends the notices if adverse action is considered?
    

Vuplicity’s developer docs include ATS and marketplace onboarding examples that map screening state into your own stages.

[Vuplicity’s examples](https://www.vuplicity.com/developers#quickstart) Fulfilled by Vuplicity

## Packages and prices

Each check is priced by the consumer reporting agency that prepares it. These are Vuplicity’s published package prices, read from its public catalog when this site was built.

Illustration Published catalog

Feed receipt Reset receipt

The receipt is fully fed. The package prices and searches are in the adjacent table.

Priced by the CRA that runs the check.

Vuplicity Catalog snapshot

Vuplicity’s published package prices
| Package | What it searches | Published price |
| --- | --- | --- |
| Basic | National criminal database, Sex offender registry, Global watchlist, Address history | $7.99 |
| Essential | All Basic checks, County and state court searches (upfront) | $14.99 |
| Complete | All Essential checks, Federal criminal search | $24.99 |

Package prices are not an all-in quote. Court and data-access fees can vary by jurisdiction.

Read from Vuplicity’s public catalog on September 29, 2026.

In the package price

The searches listed above.

Not in the package price

Court and data-access fees. Vuplicity finalizes them once the check type and the jurisdictions in the candidate’s seven-year address history are known.

Read from Vuplicity’s public catalog on Sep 29, 2026, 19:53 UTC. Fulfilled by Vuplicity [See Vuplicity’s pricing](https://www.vuplicity.com/pricing)

## Guidance, not legal advice

Hiring rules change by state, county and city. Nothing on this site is legal advice, and legal counsel should review your hiring policy decisions.

[The FTC’s guidance for employers](https://www.ftc.gov/business-guidance/resources/background-checks-what-employers-need-know)

## See it with your own AI agents.

Tell us how your product hires, or go straight to the technical details.

[Talk to Utlyze](https://www.utlyze.com/?book=call) [Explore the integration](https://yemez.com/connect)

---

# Connect your AI agent: MCP catalog, REST orders | Yemez

URL: https://yemez.com/connect

Sandbox mode. Examples only. Everything on this page is fulfilled by Vuplicity.

1.  [Home](https://yemez.com/)
2.  [Connect](https://yemez.com/connect)

For the engineer wiring it up

# Read the catalog. Order in sandbox. Get the status back.

Today your AI agent works with Vuplicity’s live endpoints. This page shows the whole path, with code to copy. Your server makes the calls, and the candidate’s details stay in the hosted flow.

-   MCP server: live, read-only
-   Orders: REST API with an organization key
-   Consent: hosted link on the candidate’s phone
-   Sandbox first, live after approval
-   Fulfilled by Vuplicity

[Start in sandbox](https://yemez.com/connect#order) [Talk to Utlyze](https://www.utlyze.com/?book=call)

   ![](/images/connect/terminal.webp)

YOUR SERVER Sandbox

1.  GET · Read the catalog `/api/v1/public/products/catalog`
2.  POST · mode: "sandbox" `/api/v1/customer/background-checks`
3.  Response · status: "created" `screening_session_url`
4.  Open the returned link yourself `candidate_email_status: "suppressed"`

Illustration · Fulfilled by Vuplicity

Press the sandbox key to replay. The candidate gives consent in the hosted flow.

AI agent stops here

Two connections. Different jobs.

## Read with MCP.  
Order with REST.

MCP reads Vuplicity’s catalog. Your server orders through REST with an organization API key.

Illustration. Toggle a cable to see its connection. Nothing is sent or saved.

   ![](/images/connect/hub.webp)    ![](/images/connect/hub-mcp.webp)    ![](/images/connect/hub-rest.webp)

Illustration · MCP reads. REST orders. Fulfilled by Vuplicity.

Read-only · No key

## Read the catalog

Point any MCP client at Vuplicity’s MCP server, or fetch the catalog as plain JSON. No credentials. The catalog can’t place orders. Its prices and availability take precedence over saved copies.

MCP (Model Context Protocol) is a common way for an AI agent to discover and call tools.

Fulfilled by Vuplicity

MCP client config

```
{
  "mcpServers": {
    "vuplicity": { "url": "https://api.vuplicity.com/api/v1/mcp" }
  }
}
```

Plain JSON

```
curl https://api.vuplicity.com/api/v1/public/products/catalog
```

Orders · Key on your server

## Order a check in sandbox

Create a Vuplicity account and accept its Terms and Consumer Reporting Certification. Then create an organization API key in Settings, under Integrations. Keep the key in server-side configuration, never in a prompt, and send it in the `x-api-key` header.

Live starts also need `employer_context.confirm_live: true` and an `idempotency_key`.

[Create a Vuplicity account](https://hr.vuplicity.com/signup) Fulfilled by Vuplicity

Start a sandbox check

```
curl -X POST https://api.vuplicity.com/api/v1/customer/background-checks \
  -H "x-api-key: $VUPLICITY_API_KEY" \
  -H "content-type: application/json" \
  -d '{
    "mode": "sandbox",
    "candidateEmail": "casey@example.com",
    "candidateFirstName": "Casey",
    "candidateLastName": "Smith",
    "checkType": "essential"
  }'
```

Sandbox first · Live after approval

## Test every state before going live

New organization accounts start in sandbox. Sandbox access unlocks once you accept the developer responsibilities, and live access follows a manual production review. Sandbox skips live data providers and live billing, and plays set scenarios.

[Sandbox docs](https://www.vuplicity.com/developers#sandbox) Fulfilled by Vuplicity

Sandbox scenario headers

```
# Sandbox scenarios for hosted sessions and direct consent requests
X-Vuplicity-Mode: sandbox
X-Vuplicity-Scenario: clear | record_found | need_info | dispute
```

## The life of one sandbox check

Illustrative sandbox sequence, fulfilled by Vuplicity

1.  Step 1 Who: Your server What happens: Sends `POST /api/v1/customer/background-checks` with `"mode": "sandbox"` and `"checkType": "essential"`. The response includes status `"created"` and a `screening_session_url`. Where it runs: Vuplicity REST API What your product can show: “Background check ordered”
2.  Step 2 Who: Vuplicity What happens: Creates the hosted consent session. In sandbox, `candidate_email_status` is `"suppressed"`, so you open the `screening_session_url` yourself. Where it runs: Vuplicity What your product can show: “Waiting on Casey’s consent” `PENDING_CONSENT`
3.  Step 3 Who: The candidate (you, in sandbox) (a person acts here) What happens: Opens the link, reviews the disclosure and gives consent. Sensitive details go into the hosted flow. AI agent stops here Where it runs: Vuplicity’s hosted flow What your product can show: Still waiting until consent is in
4.  Step 4 Who: Vuplicity What happens: Checks the order against the package and the rules for where the candidate lives and works. Where it runs: Vuplicity What your product can show: “Background check in progress” `PENDING`
5.  Step 5 Who: Vuplicity What happens: Searches the sources. Where it runs: Vuplicity What your product can show: “Background check in progress” `PROCESSING`
6.  Step 5a Who: Branch: `need_info` What happens: The CRA needs more from the candidate. The check waits in the hosted flow. Waits on a person Where it runs: Vuplicity’s hosted flow What your product can show: “Casey needs to add information” `NEED_INFO`
7.  Step 6 Who: Vuplicity What happens: The report is complete. Webhook event `background_check.completed`. Where it runs: Vuplicity What your product can show: “Report ready for review” `COMPLETE`
8.  Step 6a Who: Branch: `record_found` What happens: The report lists records for the employer to review. Where it runs: Vuplicity What your product can show: The same “Report ready for review” label
9.  Step 6b Who: Branch: `dispute` What happens: Sandbox can play a dispute, so you can design that state before it happens live. Where it runs: Vuplicity What your product can show: A state that says the CRA is reviewing it
10.  Step 6c Who: Branch: `failed` What happens: Webhook event `background_check.failed`. Move it into exception handling. Where it runs: Vuplicity What your product can show: “Needs attention” `FAILED`
11.  Step 7 Who: The employer (a person acts here) What happens: Reads the report in Vuplicity’s portal and makes the hiring decision. AI agent stops here Where it runs: Vuplicity’s portal What your product can show: The hiring team records its own decision

Field names as shown in Vuplicity’s public examples. Check the API reference before you rely on them.

## Get the status back

Subscribe to webhooks and map each event into your product. Verify the signature before you trust an event.

Webhook events for background checks
| Event | What it means (Vuplicity’s docs) | What your product can show |
| --- | --- | --- |
| `background_check.updated` | General progress update for a check in flight. | Move the stage, for example to “In progress” or “Waiting on the candidate”. |
| `background_check.completed` | The report is complete and the report URL is available. | “Report ready for review”, for the people the employer allows. |
| `background_check.failed` | The check failed and should move into exception handling. | “Needs attention”, routed to your team. |

Every delivery carries `X-Vuplicity-Signature`, `X-Vuplicity-Timestamp`, `X-Vuplicity-Event` and `X-Vuplicity-Delivery` headers.

Status names: `PENDING_CONSENT`, `PENDING`, `PROCESSING`, `NEED_INFO`, `COMPLETE`, `FAILED`.

[Webhook lifecycle docs](https://www.vuplicity.com/developers#webhook-lifecycle) Fulfilled by Vuplicity

## Paste to your AI agent

Give the AI agent that helps you build this a short, accurate brief. It points at this site’s llms.txt and Vuplicity’s endpoints, and it repeats the rules.

Prompt for your AI agent

```
You are helping us add background checks to our hiring product.
Read https://yemez.com/llms.txt first.
Read packages and prices from Vuplicity's read-only catalog: MCP server https://api.vuplicity.com/api/v1/mcp or JSON at https://api.vuplicity.com/api/v1/public/products/catalog.
Place orders only from our server with POST https://api.vuplicity.com/api/v1/customer/background-checks. The API key lives in server configuration. Never put it in a prompt or ask for it in chat.
Use "mode": "sandbox" until our account has production approval.
Never collect a candidate's sensitive details in chat. The candidate uses the hosted consent link.
Never make or imply a hiring decision. Report status to the employer.
Quote prices from the live catalog and say that package prices are not an all-in quote.
```

## Rules your AI agent follows

Vuplicity publishes these rules for AI agents that order its checks.

Illustration The rules, pinned where the work happens.

1.  **Order a check only when the employer has a permissible purpose, such as employment screening.**
    
2.  **Don’t collect the candidate’s sensitive details in chat.** Send the hosted consent link instead. It handles the disclosure and authorization.
    
    AI agent stops here
    
3.  **Don’t make or imply a hiring decision.** Report results to the employer. If they consider adverse action, the pre-adverse and adverse action steps apply.
    
    AI agent stops here
    
4.  **Quote prices from the live catalog, not from memory.** Package prices aren’t all-in. Court and data-access fees can vary by jurisdiction.
    
5.  **Treat guidance as guidance, not legal advice.** Vuplicity is not legal counsel.
    

## Before your first live check

1.  Create a developer account through Vuplicity’s organization signup.
    
2.  Accept Vuplicity’s Terms and Consumer Reporting Certification during signup.
    
3.  Create an API key in Settings, under Integrations, as an admin. Keep it on your server.
    
4.  Build in sandbox mode while live access waits for approval.
    
5.  Subscribe to webhooks and verify each signature.
    
6.  Send live starts with `employer_context.confirm_live: true` and an `idempotency_key`.
    
7.  Live access follows Vuplicity’s manual production review.
    

## Endpoints and files

Everything in this list is fulfilled by Vuplicity. Yemez’s own summaries for AI agents are at [/llms.txt](https://yemez.com/llms.txt) and [/llms-full.txt](https://yemez.com/llms-full.txt).

-   ### MCP server
    
    Fulfilled by Vuplicity
    
    `https://api.vuplicity.com/api/v1/mcp`
    
    Read-only catalog. No credentials.
    
-   ### Public catalog (JSON)
    
    Fulfilled by Vuplicity
    
    `https://api.vuplicity.com/api/v1/public/products/catalog`
    
    Read-only. No API key.
    
-   ### Order API
    
    Fulfilled by Vuplicity
    
    `POST https://api.vuplicity.com/api/v1/customer/background-checks`
    
    Organization API key in the x-api-key header. Mode: sandbox or live.
    
-   ### Developer docs
    
    Fulfilled by Vuplicity
    
    [https://www.vuplicity.com/developers](https://www.vuplicity.com/developers)
    
    Public.
    
-   ### llms.txt
    
    Fulfilled by Vuplicity
    
    [https://www.vuplicity.com/llms.txt](https://www.vuplicity.com/llms.txt)
    
    Prices, endpoints and rules in plain text.
    
-   ### OpenAPI catalog
    
    Fulfilled by Vuplicity
    
    [https://www.vuplicity.com/.well-known/vuplicity-public-catalog.openapi.json](https://www.vuplicity.com/.well-known/vuplicity-public-catalog.openapi.json)
    
    Public catalog specification.
    
-   ### Hosted screening sessions
    
    Fulfilled by Vuplicity
    
    [https://www.vuplicity.com/developers#hosted-screening-sessions](https://www.vuplicity.com/developers#hosted-screening-sessions)
    
    Vuplicity’s docs for apps that redirect the candidate from their own UI.
    

## Stuck, or ready for live checks?

Utlyze runs Yemez and takes every call, including integration questions.

[Talk to Utlyze](https://www.utlyze.com/?book=call)

---

# For CRAs: put your agency in front of AI agents | Yemez

URL: https://yemez.com/for-cras

1.  [Home](https://yemez.com/)
2.  [For CRAs](https://yemez.com/for-cras)

# Put your agency in front of AI agents.

Employers are handing hiring work to AI agents. If you run a consumer reporting agency, talk with us about making your packages something AI agents can read, price and order, with consent and hiring decisions left to people.

[Start a partner conversation](https://www.utlyze.com/?book=call)

Illustration CRA connection desk

Plug in the example Close the example

Illustrative order ticket

**AI agent orders directly from the CRA.** The CRA prepares and delivers the report to the employer. AI agent stops here Consent and hiring decisions stay with people.

Illustration of a partner conversation. Yemez publishes connection details, rules and a catalog. Today, AI agents order from Vuplicity directly.

## How it works today, with Vuplicity

Illustrative

Yemez: connection details, rules, catalog read

reads

1.  The employer’s AI agent
2.  orders with the employer’s key
3.  The CRA: Vuplicity
4.  report
5.  The employer

Illustrative. The AI agent orders from the CRA directly. The CRA prepares the report and delivers it to the employer. Yemez does not receive, prepare or furnish reports.

## How it would fit

-   ### You stay the CRA.
    
    Your agency prepares and delivers every report. Yemez is not a consumer reporting agency and does not prepare or furnish reports.
    
-   ### You set the price.
    
    Checks are priced by the CRA that prepares them, and AI agents quote the prices you publish.
    
-   ### Consent and decisions stay with people.
    
    The candidate gives consent through your hosted link on their phone. Hiring decisions stay with the employer.
    

## Where things stand

-   Live today
    
    Vuplicity prepares and delivers the checks that AI agents order using the connection details on this site.
    
-   Open
    
    Conversations with other consumer reporting agencies about the same pattern.
    

## What a partner conversation covers

-   Which of your packages AI agents should be able to order
-   How AI agents reach your agency, and how candidates get your consent link
-   How status and reports get back to the employer
-   What you need from us, and when

## Start the conversation.

Tell us about your agency and the employers you serve.

[Start a partner conversation](https://www.utlyze.com/?book=call)

---

# FAQ: AI agents, background checks and consent | Yemez

URL: https://yemez.com/faq

1.  [Home](https://yemez.com/)
2.  [FAQ](https://yemez.com/faq)

# Questions, answered plainly

Who prepares the check, what it costs, how the candidate gives consent, and why the hiring decision stays with the employer.

## About Yemez

### What is Yemez?

Yemez is the background check platform built for AI agents. It shows AI agents how to order background checks from a consumer reporting agency (CRA) and points them to that CRA’s live connection details, so the AI agents in your hiring product can add the background check step.

### Is Yemez a consumer reporting agency?

No. Yemez does not prepare or furnish consumer reports. Every check is prepared and delivered by a consumer reporting agency. Today, that is Vuplicity.

### How are Yemez and Vuplicity related?

They are separate brands. Vuplicity is a consumer reporting agency, and today it prepares and delivers the checks described on this site. Utlyze runs Yemez.

### Who runs Yemez?

Utlyze. Every call booked from this site is with the Utlyze team.

## Checks and prices

### What does a background check cost?

Checks are priced by the CRA that prepares them. Vuplicity’s published package prices, read from its public catalog on September 29, 2026: Basic $7.99, Essential $14.99 and Complete $24.99. Package prices are not an all-in quote. Court and data-access fees can vary by jurisdiction.

### Why isn’t the package price the total?

Court and data-access fees depend on where the search runs. Vuplicity finalizes them once the check type and the jurisdictions in the candidate’s seven-year address history are known.

### What is in each package?

Basic: National criminal database, Sex offender registry, Global watchlist, Address history. Essential: All Basic checks, County and state court searches (upfront). Complete: All Essential checks, Federal criminal search.

## Consent and decisions

### How does the candidate give consent?

Through a secure link on their phone, hosted by the CRA. They review the disclosure and give consent there. Their sensitive details go into that hosted flow, not into your AI agent’s chat.

### Can our AI agent ask the candidate for their details in chat?

No. Vuplicity’s rules for AI agents say not to collect the candidate’s sensitive details in chat. The AI agent sends the hosted consent link instead.

### Who decides who gets hired?

The employer. Your AI agent orders the check and reports the status. The employer reads the report and makes the hiring decision.

### What if the employer considers adverse action?

The pre-adverse and adverse action steps apply. The FTC explains them in [its guidance for employers](https://www.ftc.gov/business-guidance/resources/background-checks-what-employers-need-know).

### Is anything on this site legal advice?

No. Hiring rules change by state, county and city, and legal counsel should review your hiring policy decisions.

## Building the integration

### Can our AI agent order through MCP?

Not today. Vuplicity’s MCP server is a read-only catalog, so your AI agent can read packages and prices there. Orders go through the REST API with an organization API key kept on your server.

### Can we try it before running live checks?

Yes. New Vuplicity organization accounts start in sandbox mode, which skips live data providers and live billing. Live access comes after a manual production review.

### Where should the API key live?

In server-side configuration. Never in a prompt, and never in a chat.

### Is there an SDK?

Vuplicity publishes a TypeScript SDK, and its hosted-session update is waiting on a verified release. Until then, the Connect page uses plain REST examples.

### Can someone order a check from a general AI assistant today?

Not through an assistant app store today. An AI assistant that can use MCP can read the public catalog. Orders go through the REST API with an organization key.

### Do we need to be technical?

Someone on your team needs to call a REST API from a server. If that isn’t you, [talk to Utlyze](https://www.utlyze.com/?book=call) and we’ll help you work out the next step.

### I run a CRA. Can AI agents order from us?

Talk with us. Start a partner conversation from [the For CRAs page](https://yemez.com/for-cras).

### What does yemez.com collect about me?

Almost nothing. The site sets no cookies, runs no analytics and has no forms. Read [the privacy policy](https://yemez.com/privacy).

## Still have a question?

Ask a person.

[Talk to Utlyze](https://www.utlyze.com/?book=call)

---

# Blog: AI agents and background checks | Yemez

URL: https://yemez.com/blog

1.  [Home](https://yemez.com/)
2.  [Blog](https://yemez.com/blog)

# Plain answers about AI agents and background checks

For teams building AI hiring software and the people who approve it. Every legal statement links its primary source. Guidance, not legal advice.

[Follow by RSS](https://yemez.com/blog/rss.xml)

## Add the background check to your hiring product.

[Talk to Utlyze](https://www.utlyze.com/?book=call)