HrFlow.ai updates
HrFlow.ai updates
hrflow.ai

Introducing storage duration management​

 

New

  

We are happy to announce that it is now possible for our customers to choose the storage_duration for profile attachments (profile picture, cover letter, resume) to comply with their internal privacy and security policies.

When you create a source on HrFlow.ai, you can set the storage_duration value to: never_store, 2H, 1D, 1W, 1M, 6M, 1Y, 2Y, or never_expire.

Note that this value cannot be changed after creation.

Screenshot 2022-12-13 at 12.00.17.png

Improving multi-lingual Date Parsing

 

Improvement

  

We have improved the detection of dates across all our certified languages.

The impact can be most noticed on German resumes.

Extending Parsing to octet-stream files

 

Fix

  

We are pleased to announce that octet-stream files without extensions can now be successfully parsed.

This fix improves the overall performance of our Parsing and ensures that our system correctly handles these rare exceptional files.

Introducing the Tagging API, our AI library for Talent & Workforce data classification

 

New

  

In response to the increasing demand from customers who want to categorize and classify HR data (texts, documents, sections, paragraphs) automatically, we are thrilled to announce our latest AI module: Tagging API.

The Tagging API is a library of pre-built AI algorithms that automatically normalize and classify talent and workforce data, including job titles, skills, and more.

Our first release includes 4 AI algorithms, called taggers, trained to make 4 different types of predictions:

  • tagger-rome-family: predicts the job family (Grand Domaine) from any raw input text such as a resume, job offer, job section, or experience, based on the job classification open-sourced by the french government, called "ROME".
  • tagger-rome-subfamily: predicts the job subfamily (Domaine) from any raw input text
  • tagger-rome-category: predicts the job category (Fiches Métiers) from any raw input text
  • tagger-rome-jobtitle: predicts the job title (Intitulé de poste) from any raw input text
  • revealing: predicts hard and soft skills from any raw input text based on the HrFlow.ai internal knowledge graph.

The JSON output of the API provides the list of the most likely predictions with associated names, IDs, and probabilities.

In the coming month, our team will release more AI algorithms to cover a broader range of use cases.

To integrate this new API, you can check the following documentation:

Cegid-Talentsoft Connector

 

New

  

About Cegid-Talentsoft :

Talentsoft (by Cegid) offers a full-suite HCM solution so you can keep all your HR processes in one platform.

  • Make HR processes easy & simple for your teams.
  • Help HR & IT leaders transform organizations.
  • Accelerate HCM productivity for better performance.

Connector features:

  • Pull profiles: Retrieves profiles from Cegid Talentsoft candidates export API and sends them to a Hrflow.ai Source.
  • Pull jobs: Retrieves jobs from Cegid Talentsoft vacancies export API and sends them to a HrFlow.ai Board.
  • Catch profile : Receives profiles. It catches three different events :
    • applicant_new event by fetching every newly created profile from TalentSoft and sending it to HrFlow.ai Parsing API.
    • applicant_update event by only updating the profile's structured data coming from TalentSoft in HrFlow.ai.
    • applicant_resume_update event by running a new HrFlow.ai Parsing on the updated resume.

Useful links:

📄 Visit the Cegid talentsoft to learn more

💻 Connector code on our GitHub.

TS - Dataflow.png

TalentSoft - Profiles Listing.png

SmartRecruiters Connector

 

New

  

About SmartRecruiters:

SmartRecruiters’ Talent Acquisition Suite is used by organizations to make the best hires. It has complete functionality for recruitment marketing and collaborative hiring built on a modern cloud platform with an open marketplace for 3rd party recruitment services.

Connector features:

SmartRecruiters integration with HrFlow allows you to :

  • Pull jobs: Retrieves all jobs from SmartRecruiters and sends them to an HrFlow board.

  • Pull profiles: Retrieves profiles from SmartRecruiters and sends them to an Hrflow.ai Source.

  • Push profiles: Send profiles from an HrFlow source to SmartRecruiters as a candidate for a given job

Useful links:

📄 Visit the Smartrecruiters API documentation to learn more

💻 Connector code on our GitHub.

Uploading image 50%

HrFlow x SmartRecruiters Connector (Data Flow).png

Connect SmartRecruiters to +200 data sources.png

SmartRecruiters - Profiles listings.png

HrFlow.ai Parsing v4.0.0 Updates

 

Improvement

 

New

 

Fix

  

We are excited to announce that HrFlow.ai Parsing V4.0 is finally available. This release includes awesome new features and improvements that we hope you enjoy.

State-of-the-art Recognition.png

Improvements

  • Complex layout parsing for all our 11 certified languages (lien)
  • Profile picture detection: small size image
  • Gender detection: taking into account the writing style in addition to the profile picture
  • Section classification: 97% accuracy (vs 92% in V3.0)
  • Special char polices management
  • Location detection
  • Scan orientation detection

New Features

  • Company logo detection
  • URLs detection
  • Phone number detection with E.164 standard (e.g. 06100000 + France ➜ +336100000)

These updates are backward compatible. Hence, they are no API migration that needs to be taken, except for the company logo detection feature.

To check the complete Profile JSON, please head over to the Profile Object page: https://developers.hrflow.ai/reference/the-profile-object .

To give a try to our latest Parsing version, please visit:

Profile Parsing v3.3.24 Updates

 

Improvement

  

The Parsing API developed by HrFlow.ai has been updated. Version 3.3.24 ships with an updated URLs schema. Please find below an example of the “URLs” field content:

"urls": [
  {"type": "from_resume", "url" : "https://hrflow.ai/"},
  {"type": "linkedin", "url" : "https://fr.linkedin.com/doe-john"},
  {"type": "twitter", "url" : "https://twitter.com/hrflowai"},
  {"type": "facebook", "url" : "https://www.facebook.com/hrflowai/"},  
  {"type": "github", "url" : "https://github.com/harry-james-potter"}, 
]

Further information about our Parsing API can be found at: developers.hrflow.ai

Feel free to test our Parsing update through: labs.hrflow.ai

Profile Parsing v3.3.23 Updates

 

Fix

 

Improvement

  

The Parsing API developed by HrFlow.ai has been updated. Version 3.3.23 ships with a better date detection algorithm.

Further information about our Parsing API can be found at developers.hrflow.ai

Feel free to test our new profile parsing potential through labs.hrflow.ai

Profile Parsing v3.3.22 Updates

 

Improvement

  

The Parsing API developed by HrFlow.ai has been updated. Version 3.3.22 ships with an enhanced experience detection algorithm.

Further information about our Parsing API can be found at developers.hrflow.ai

Feel free to test our new profile parsing potential through labs.hrflow.ai