View source for Integration Documentation: Affinda AI-Powered Resume Parser

== Introduction ==
This document introduces the integration of an AI-powered Resume Parser powered by Affinda.com, a platform that employs the OpenAI (ChatGPT) engine for enhanced OCR and parsing results. This integration aims to streamline the onboarding process for candidates by minimizing manual data entry.
----

== Setup & Configuration ==
Step-by-step Integration Setup:

# Account Creation: Begin by creating an account with Affinda to obtain the API credentials.
# Affinda Settings:
#* Navigate to <code>Setup</code> -> <code>Custom Settings</code> -> <code>Affinda Settings</code>.
#* Fill in the fields "API Key" and either "Collection Id" or "Workspace Id".

----

== Utilizing the Out-of-the-box <code>ResumeParser</code> Page ==
For a straightforward implementation, use the built-in <code>b3o__ResumeParser</code> Visualforce page:

# Access the "Document Pack" tab.
# Add a new Document and select "External Link" as the document type.
# Populate the "External Link" field with <code>/b3o__ResumeParser</code>.

By doing so, the resume parser becomes a part of the document packs assigned to candidates.
----

== Customizing the <code><resume-parser></code> Component ==
Developers have the flexibility to embed the <code>ResumeParser</code> page in various ways, including within 3B Portals, adding extra logic, or updating styling:

Instructions:

* Embed the web component scripts
<syntaxhighlight lang="html">
<script src="{!URLFOR($Resource.b3o__GeneralUtils, 'index.js')}" defer="true" type="module"></script>
<script src="{!URLFOR($Resource.b3o__ResumeParser, 'index.js')}" defer="true" type="module"></script>

</syntaxhighlight>

* Add the <code><resume-parser></code> component to the page and initialize with the globals object:
<syntaxhighlight lang="html">
<resume-parser 
    user-id="{!$CurrentPage.parameters.conId}"
    workspace="{!$Setup.b3o__Affinda_Settings__c.b3o__Workspace_Id__c}"
    collection="{!$Setup.b3o__Affinda_Settings__c.b3o__Collection_Id__c}"
    api-key="{!$Setup.b3o__Affinda_Settings__c.b3o__API_Key__c}"
></resume-parser>

<script>
    const globals = {
        sessionId: '{!GETSESSIONID()}',
        siteUrl: '{!$Site.baseUrl}',
        useREST: false,
        router: b3o.GlobalRemotingRouter
    }
    window['globals'] = globals;
</script>

</syntaxhighlight>Component Attributes:

* <code>user-id</code>: Salesforce contact id
* <code>workspace</code>: Affinda workspace id
* <code>collection</code>: Affinda collection id
* <code>api-key</code>: Affinda API key
* <code>success-message</code>: Message displayed upon successful parsing

== Data Management by the Resume Parser ==
When a CV/Resume is parsed, various Salesforce records and fields are updated:

Contact:

* b3o__Linkedin_Profile_Link__c
* b3o__Profession__c
* b3o__Summary__c
* BirthDate
* Address & Coordinates

b3o__Language__c:

* Name

b3o__Certification__c:

* Name

b3o__Skill__c:

* Name
* b3o__Type__c (Options: Soft/Hard Skill)

b3o__Employment_History__c:

* b3o__Start_Date__c
* b3o__End_Date__c
* b3o__Grade__c
* b3o__Institution_Name__c
* b3o__Degree_Program__c

b3o__Education_History__c:

* b3o__Start_Date__c
* b3o__End_Date__c
* b3o__Grade__c
* b3o__Institution_Name__c
* b3o__Degree_Program__c

b3o__Reference__c:

* b3o__First_Name_Referee__c
* b3o__Last_Name_Referee__c
* b3o__Email_Referee__c
* b3o__Phone_Referee__c
* b3o__Position_Referee__c

----

== Considerations ==

* Accuracy: Resume parsing, while technologically advanced, may not always be 100% accurate. Affinda's engine is among the top-tier due to its usage of OpenAI (ChatGPT) but discrepancies might still occur.
* Candidate Review: Admins should consider integrating a 3B Form to allow candidates the opportunity to review and adjust parsed results. The primary goal is to reduce onboarding time and manual input, not to entirely replace human review.

----

== Additional Tips ==

* Version Compatibility: Regularly review updates from both Salesforce and Affinda to ensure compatibility and leverage new features.
* User Feedback: Encourage feedback from end-users to continuously improve the implementation and adapt to specific organizational needs.

[[Category:3B Onboarding]]
[[Category:3B Portals]]