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

# September 2026 Release Notes

|  Release Name  | Dataset Version | Publish Date |
| :------------: | :-------------: | :----------: |
| September 2026 |      v35.2      |  09/15/2026  |

**Welcome to our September 2026 release notes!**

Back to school season is here, and we've got an exciting batch of updates to share for show-and-tell this month: improved salary and compensation data, a new sandbox endpoint for the Job Posting API, improved deduplication, and expanded Job Posting and Technology Data coverage.

Check it out:

* 💰 **[Inferred Salary Improvements](#inferred-salary-improvements)**: Expanded and improved salary data, powered by more real-world signals from our job posting dataset.
* ▶️ **[Job Posting Sandbox API](#job-posting-sandbox-api)**: A free way to test our Job Posting API using a test dataset of curated job posts.
* 📈 [**Data Quality & Coverage Improvements**](#improvements-and-bug-fixes): 65% less duplication of LinkedIn profiles, validation of GitHub URLs, and expanded coverage of Job Posting and Technology Data.
* **📊 [Freshness Updates](#freshness)**: This past month, we refreshed [57M+ resume experiences](#freshness) and [detected 6.6M+ job changes](#job-changes) across our global dataset

***

## ✨ Release Updates

### Inferred Salary Improvements

<Badge color="blue">Person Data</Badge> <Badge color="green">GA</Badge> <Badge color="yellow">✨ Product Update</Badge> <Badge color="silver">📐 Schema Change</Badge>

**What's New?**

We've improved how our `inferred_salary` field is calculated and added three new fields to provide a more precise and transparent view of estimated compensation.

Our updated methodology now **incorporates compensation signals from real-world job posts**, accounting for factors such as role, seniority, company, geography, and recency.

Alongside the existing `inferred_salary` field, we have also added the following fields:

**New Inferred Salary Fields**

| FIELD NAME                                                              | FIELD TYPE      | FIELD BUNDLE                                                       | FIELD DESCRIPTION                                                                                                        |
| :---------------------------------------------------------------------- | :-------------- | :----------------------------------------------------------------- | :----------------------------------------------------------------------------------------------------------------------- |
| [`inferred_salary_low`](/docs/fields#inferred_salary_low)               | `Integer`       | [Inferred Salary](/docs/person-data-field-bundles#inferred-salary) | The lower bound (tenth percentile) of the distribution of salaries associated with this role.                            |
| [`inferred_salary_high`](/docs/fields#inferred_salary_high)             | `Integer`       | [Inferred Salary](/docs/person-data-field-bundles#inferred-salary) | The upper bound (90th percentile) of the distribution of salaries associated with this role.                             |
| [`inferred_salary_confidence`](/docs/fields#inferred_salary_confidence) | `String (Enum)` | [Inferred Salary](/docs/person-data-field-bundles#inferred-salary) | The confidence tier associated with the inferred salary. Will be one of the Canonical Inferred Salary Confidence Values. |

These fields offer a more granular estimated range and additional context about the strength of each salary estimate.

<Note>
  The [canonical values](/docs/inferred-salaries#inferred-salaries) for the `inferred_salary` field are not changing. We are only improving the methodology used to assign those canonical salary buckets to each record.
</Note>

**Example**

```json theme={null}
{
  "inferred_salary": "150,000-250,000",
  "inferred_salary_low": 150000,
  "inferred_salary_high": 210000,
  "inferred_salary_confidence": "high"
}
```

**Why It Matters**

These improvements make Inferred Salary data more useful across a wider range of compensation, targeting, and analytics workflows.

With these updates, you can expect:

* More accurate salary estimates across international markets
* Greater granularity for compensation above \$250k
* Broader coverage
* More precise estimated salary ranges
* A confidence indicator to help evaluate each estimate

By incorporating real-world compensation signals from our Job Posting Dataset, our salary estimates now better reflect current market compensation across roles, companies and geographies.

**Availability**

The new fields are included in the premium [Inferred Salary field bundle](/docs/person-data-field-bundles#inferred-salary), alongside the existing `inferred_salary` field.

Customers who already have access to this bundle will automatically receive the new fields in their upcoming deliveries.

If you're an enterprise customer and would like access to these fields, please reach out to your Customer Success team. Self-serve customers can add the bundle to their subscription through the API Dashboard under **Plans & Billing** > **Subscriptions** > **Manage**.

**Additional Resources**

* [Field Bundle](/docs/person-data-field-bundles#inferred-salary)
* [Schema Definition](/docs/fields#inferred_salary)
* [Fill Rate Stats](/docs/resume-stats)

***

### Job Posting Sandbox API

<Badge color="blue">Job Posting Data</Badge> <Badge color="green">GA</Badge> <Badge color="yellow">✨ Product Update</Badge>

**What's New?**

This month, we released a new [Sandbox endpoint](/docs/sandbox-apis) for our **Job Posting Search API**; you can now **test and build with the Job Posting Search API for free** using our new Sandbox endpoint.

The Sandbox Job Posting Search API mirrors the functionality of the production API, but runs against a sample of real Job Posting data and does not consume credits.

This offers a risk-free environment to test queries, build integrations, and prototype workflows before moving to production.

**Overview**

Endpoint: [`https://sandbox.api.peopledatalabs.com/v5/job_posting/search`](https://sandbox.api.peopledatalabs.com/v5/job_posting/search)

The Sandbox endpoint supports the same input parameters and returns the same JSON response structure as the production Job Posting Search API.

The underlying Sandbox dataset contains a sample of **5,000 real Job Posting records** and is **refreshed monthly**, providing enough data to test workflows including filtering, pagination, and aggregations.

**Why It Matters**

Previously, testing the Job Posting Search API required access to production credits. With the new Sandbox endpoint, you can:

* Build and test integrations and workflows without consuming credits
* Experiment with search queries before using them in production
* Test filtering, pagination, and aggregation workflows against real Job Posting Data
* Evaluate Job Posting Data without needing test credits provisioned first

Because the Sandbox uses real Job Posting records, you can also evaluate the structure of the underlying data using representative samples.

<Warning>
  **Sandbox Dataset Limitations**

  The Sandbox endpoint uses a small curated dataset of 5,000 job posts and is only updated monthly.

  Therefore, the Sandbox API is **not reflective of the coverage and freshness of the production Job Posting Search API**. Instead, this endpoint is intended to provide a free way to test functionality and workflows using the endpoint.

  If you would like to evaluate the production Job Posting Search API, please [reach out](https://www.peopledatalabs.com/job-posting-beta-sign-up) to us.
</Warning>

**Availability**

The Sandbox endpoint for our Job Posting Search API is [available across all plans and does not consume API credits](/docs/sandbox-apis-reference#billing-and-access). To access the Sandbox API, simply query the endpoint with your PDL API key.

If you do not have an API key, you can [sign up](https://dashboard.peopledatalabs.com/) for a free one through our self-serve portal.

**Getting Started**

To immediately start sending test queries, check out this quick copy-and-paste code example: [Sandbox Example](/docs/sandbox-apis-examples#sandbox-job-posting-search-api)

For more test queries to try out, you can also run any of the code samples in the [Job Posting Search API Examples](/docs/examples-job-posting-search-api) page. Simply replace the production Job Posting Search API endpoint in those examples with the sandbox endpoint:

**To run any [Job Posting Search API example](/docs/examples-job-posting-search-api) with the Sandbox Endpoint:**

| Replace                                                |
| :----------------------------------------------------- |
| `https://api.peopledatalabs.com/v5/job_posting/search` |

| With                                                           |
| :------------------------------------------------------------- |
| `https://sandbox.api.peopledatalabs.com/v5/job_posting/search` |

***

## 🚀 Data Updates

### Freshness

*The number of resume experiences and locations verified in our datasets (based on the [`job_last_verified`](/docs/fields#job_last_verified) and [`location_last_updated`](/docs/fields#location_last_updated) fields).*

**Monthly (v35.1 → v35.2)**
*Freshness updates over the past month.*

| Dataset | Geography     | Field        | Records Updated |
| :------ | :------------ | :----------- | :-------------- |
| Resume  | Global        | `experience` | 57,443,384      |
| Resume  | Global        | `location`   | 61,896,943      |
| Resume  | United States | `experience` | 18,433,367      |
| Resume  | United States | `location`   | 18,171,267      |

### Job Changes

*The number of person records where the primary job experience changed in our Person Dataset (based on the [`job_last_changed`](/docs/fields#job_last_changed) field).*

**Monthly (v35.1 → v35.2)**
*Job changes detected over the past month.*

| Dataset | Geography     | Records Updated |
| :------ | :------------ | :-------------- |
| Resume  | Global        | 6,657,429       |
| Resume  | United States | 1,460,051       |

***

## 🌐 Coverage

### Monthly Highlights (v35.1 → v35.2)

<Card title="Resume Dataset" href="/docs/resume-stats">
  | Linkage                                           | Coverage in v35.1 | Coverage in v35.2 | Increase (%) |
  | :------------------------------------------------ | ----------------- | ----------------- | ------------ |
  | Total Records                                     | 814,867,031       | 815,476,877       | 0.07%        |
  | [`inferred_salary`](/docs/fields#inferred_salary) | 274,335,234       | 346,605,387       | 26.34%       |
  | [`headline`](/docs/fields#headline)               | 273,446,294       | 295,097,900       | 7.92%        |
  | [`skills`](/docs/fields#skills)                   | 205,364,885       | 216,221,372       | 5.29%        |
  | [`github_url`](/docs/fields#github_url)           | 2,075,559         | 2,182,448         | 5.15%        |
  | [`github_username`](/docs/fields#github_username) | 2,075,559         | 2,182,448         | 5.15%        |
</Card>

<Card title="Full Dataset" href="/docs/stats">
  | Linkage       | Coverage in v35.1 | Coverage in v35.2 | Increase (%) |
  | :------------ | ----------------- | ----------------- | ------------ |
  | Total Records | 2,464,042,232     | 2,424,643,098     | -1.60%       |
</Card>

<Card title="Company Dataset" href="/docs/company-stats">
  | Linkage                                                                     | Coverage in v35.1 | Coverage in v35.2 | Increase (%) |
  | :-------------------------------------------------------------------------- | ----------------- | ----------------- | ------------ |
  | Total Records                                                               | 74,672,539        | 76,606,079        | 2.59%        |
  | [`technologies_used`](/docs/company-schema#technologies_used)               | 6,462,182         | 6,758,244         | 4.58%        |
  | [`active_job_postings`](/docs/company-schema#active_job_postings)           | 109,709           | 115,918           | 5.66%        |
  | [`deactivated_job_postings`](/docs/company-schema#deactivated_job_postings) | 109,709           | 115,918           | 5.66%        |
</Card>

<Card title="Job Posting Dataset" href="/docs/jp-stats">
  | Metric                 | Coverage in v35.1 | Coverage in v35.2 | Increase (%) |
  | :--------------------- | ----------------- | ----------------- | ------------ |
  | Total Records          | 40,614,651        | 46,188,396        | 13.72%       |
  | Total Unique Companies | 109,824           | 116,514           | 6.09%        |
  | Total Active Postings  | 6,293,692         | 7,024,184         | 11.61%       |
</Card>

<Card title="Technology Data">
  <div style={{whiteSpace: "nowrap"}}>
    | Metric                                                                                                                                                                         | Coverage in v35.1 | Coverage in v35.2 | Increase (%) |
    | :----------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | ----------------- | ----------------- | ------------ |
    | <Tooltip tip="The total number of standardized technologies included in our technology taxonomy and available for identification.">Technologies Tracked</Tooltip>              | 25,566            | 25,566            | 0.00%        |
    | <Tooltip tip="The number of distinct technologies observed across all companies in the dataset.">Technologies Observed</Tooltip>                                               | 22,249            | 22,571            | 1.45%        |
    | <Tooltip tip="The number of companies for which at least one technology has been observed.">Companies with Technology Data</Tooltip>                                           | 6,462,182         | 6,758,244         | 4.58%        |
    | <Tooltip tip="The total number of unique company-to-technology relationships observed in the dataset.">Observed Company\<>Technology Pairs</Tooltip>                           | 53,635,904        | 55,632,265        | 3.72%        |
    | <Tooltip tip="The total number of times technologies are referenced across the underlying data sources used to identify technology usage.">Total Technology Mentions</Tooltip> | 186,560,858       | 197,780,611       | 6.01%        |
  </div>
</Card>

### Commentary

* **Person Data**
  * [`inferred_salary`](/docs/fields#inferred_salary) coverage in the Resume Dataset increased by **26%** following this month's [salary data improvements](#inferred-salary-improvements)
  * GitHub \<> LinkedIn profile linkages increased by **5%**, aided by this month's [GitHub URL validation](#improvements-and-bug-fixes) update
  * The [Full Dataset](/docs/stats) decreased by approximately **40M** records as a result of this month's [LinkedIn deduplication improvements](#improvements-and-bug-fixes)
* **Company Data**
  * The number of companies with [Technology Data](/docs/company-schema#technologies_used) increased by **4.5%**
  * The number of companies with [Job Posting Data](/docs/jp-stats) increased by **5.7%**
* **Job Posting Data**
  * [Job Posting Dataset](/docs/jp-stats) coverage continued to expand with **6% more unique companies** and **12% more active postings** this month
* **Technology Data**
  * Total technology mentions increased by **6%**, while the number of unique Company\<>Technology linkages increased by **3.7%**
  * We also observed Technology Data across **4.6% more companies** this month

***

## 🛠 Improvements and Bug Fixes

### Improvements

* **Person Record Deduplication**
  * We reduced the number of duplicate LinkedIn profiles by **65%** by filtering sparse records that duplicate more complete profiles. This removed approximately **40M** records from the full dataset (see [commentary above](#commentary)).
  * These improvements primarily target the full Person Dataset, and do not impact our resume or other datasets.
  * Customers using the full Person Dataset may see a decrease in the total number of records. Customers using the [person changelog](/docs/person-changelog-api) will see a temporary spike in the number of records marked as [`deleted`](/docs/output-response-person-changelog-api#datadeleted) this month.
* **GitHub URL Validation**
  * The [`github_url`](/docs/fields#github_url) field on person records now only contains URLs that return a `200` status. URLs returning a `404` are retained in the [`profiles`](/docs/fields#profiles) array, but will no longer appear as the top-level `github_url`.
* **Improved Company Location Coverage**
  * We expanded our company location sourcing to increase our coverage of locations across records in our Company Dataset.Company Location Coverage


## Related topics

- [February 2026 Release Notes](/changelog/february-2026-release-notes-v331.md)
- [September 2024 Release Notes](/changelog/september-2024-release-announcement-v272.md)
- [September 2025 Release Notes](/changelog/september-2025-release-notes-v312.md)
- [January 2026 Release Notes](/changelog/january-2026-release-notes-v330.md)
- [June 2026 Release Notes](/changelog/june-2026-release-notes-v342.md)