Method and system for managing, matching, and sourcing employment candidates in a recruitment campaign

• US 10,318,927 B2
• Filed: 05/01/2018
• Issued: 06/11/2019
• Est. Priority Date: 07/17/2017
• Status: Active Grant

• Current Assignee: ExpertHiring, LLC
• Original Assignee: ExpertHiring, LLC
• Inventors: Champaneria, Paul
• Primary Examiner(s): Whitaker, Andrew B
• Application Number:
  US15/967,971
• Publication Number:
  US 20190019160A1
• Time in Patent Office: 406 Days
• Field of Search: 705320-321
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 40/295 Named entity recognition
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06N 20/10 using kernel methods, e.g. support vector machines [SVM]
• CPC Class Codes: G06N 3/04 Architecture, e.g. interconnection topology
• CPC Class Codes: G06N 5/01 Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound
• CPC Class Codes: G06N 5/04 Inference or reasoning models
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06Q 10/1053 Employment or hiring

Abstract

A method and system for automating some aspects of a recruiting process, which may implement rules permitting the processes of sourcing candidates, setting up job interviews, and responding to candidate questions to all be automated with a computer. An automated system may specifically be used for conducting a conversation with a candidate over one or more communications channels, which may include pre-interview vetting of the candidate and clarification of aspects of the candidate'"'"'s provided information, and may include conducting an interview or some aspects thereof. It may also be understood that an automated system may, in addition to conversing with the candidate, be used to provide feedback or additional opportunities to the candidate.

Claims

1. A computer-implemented method for performing automated recruitment to a position, the method comprising:

providing an automated recruitment system, the automated recruitment system comprising a natural language processing engine and a database comprising a plurality of intents of different types and a plurality of entities of different types, the intents being natural language processing intents and the entities being natural language processing entities, the plurality of intents and the plurality of entities configured into a plurality of conversation paths likewise stored in the database, each of the plurality of conversation paths comprising an expression of at least one intent about at least one entity;

identifying at least one job description, at least one job requirement for the job description, and at least one candidate having candidate ré

sumé

information associated therewith;

initiating a conversation with the at least one candidate based on the plurality of conversation paths stored in the database;

processing first instructions and information associated with a first intent that outputs a first message to the at least one candidate;

processing second instructions and information associated with a candidate response message received from the at least one candidate, identifying one or more candidate response intents in the candidate response message, and mapping the one or more candidate response intents to a conversation path in the plurality of conversation paths;

processing third instructions associated with the conversation path in the plurality of conversation paths, and generating and providing to the at least one candidate a recruiter message based on the third instructions;

wherein the recruiter message comprises at least one of a question for the at least one candidate generated based on the candidate ré

sumé

information and a provision of further information to the at least one candidate based on an expression of candidate interest in the one or more candidate response intents.

2. The method of claim 1, wherein the first intent is a casual intent comprising a conversation opener in a first language, wherein the automated recruitment system is configured to verify that a response of the at least one candidate is in the first language, and wherein the automated recruitment system is configured to continue using the first language when the response of the at least one candidate is in the first language and is configured to identify a response language of the at least one candidate and use the response language of the at least one candidate when the response of the at least one candidate is in a second language.

3. The method of claim 1, wherein the step of identifying one or more candidate response intents comprises identifying whether the one or more candidate response intents are casual intents, business intents, or other intents, the other intents being intents not classified as casual intents or business intents, and, when an intent in the one or more candidate response intents is identified as including at least one other intent, sending a notification message to a recruiter regarding the other intent.

4. The method of claim 1, further comprising:

determining whether a candidate response message warrants immediate transfer of the conversation to a recruiter according to predefined criteria, the predefined criteria including at least one of a significant and sudden increase in audio volume from the at least one candidate during a live conversation and a manifestation from the at least one candidate that they do not wish to speak to a robot; and

immediately transferring the conversation to the recruiter when the predefined criteria are satisfied.

5. The method of claim 1, wherein the method comprises providing a plurality of instances of the automated recruitment system, each instance of the automated recruitment system being paired with a different recruiter and having separately customizable settings.

6. The method of claim 1, wherein the automated recruitment system is configured to send and receive candidate response messages using a plurality of communications systems;

and

wherein the database is configured to store ongoing conversations with candidates in conversation-specific form such that candidate response messages received via the plurality of communications systems are consolidated into a single ongoing conversation in the database.

7. The method of claim 6, wherein a first communication system in the plurality of communications systems is a text system and a second communication system in the plurality of communications systems is a speech system;

wherein the automated recruitment system further comprises a speech-to-text transcription system and a real-time text-to-speech generation system, and wherein the method further comprises;

transcribing, on the automated recruitment system, a voice conversation into a text transcription, and storing the text transcription in the single ongoing conversation;

processing instructions based on the text transcription; and

generating a voice response using the text-to-speech generation system and providing the voice response to the at least one candidate.

8. The method of claim 6, wherein the automated recruitment system is configured to identify one or more informational gaps in at least one ongoing conversation with the at least one candidate, and wherein the automated recruitment system is configured to automatically generate and proactively send a message to the at least one candidate directed to filling the one or more informational gaps, the message comprising at least one of:

a question about the level of experience of the at least one candidate, a question about the desire of the at least one candidate for a job change, or a question about the interest that the at least one candidate has in a specific position.

9. The method of claim 6, wherein the automated recruitment system is configured to generate and send messages to the at least one candidate according to an escalation hierarchy of messaging systems, such that the automated recruitment system is configured to send a first message via a first messaging system chosen from the plurality of communications methods, and such that the automated recruitment system is configured to send a second message via a second messaging system chosen from the plurality of communications systems, the second messaging system being different from the first messaging system and being positioned at a higher position on the escalation hierarchy.

10. The method of claim 1, wherein the automated recruitment system is configured to perform crawling, scraping, and parsing of searchable data, and is further configured to amend the candidate ré

sumé

information associated with the at least one candidate to incorporate at least one element of the searchable data.

11. The method of claim 10, wherein the step of crawling, scraping, and parsing of searchable data comprises at least one of:

identifying a social media presence of the at least one candidate and identifying at least one of;

a recent social media posting of the at least one candidate, a newly added skill, a newly added work experience, and a recommendation from another user;

identifying a social media presence of a previous or current employer or reference of the at least one candidate and identifying at least one social media posting of the previous or current employer or reference;

identifying a staff directory page of a previous or current employer of the at least one candidate and verifying a presence of the at least one candidate on the staff directory page;

identifying corroborating information for one or more candidate ré

sumé

information elements flagged as deficient, unknown, or potentially fraudulent by the automated recruitment system.

12. The method of claim 1, further comprising:

analyzing, with the automated recruitment system, a level of similarity between the candidate ré

sumé

information and the at least one job requirement of the job description, and identifying a deficient aspect of the candidate ré

sumé

information, the deficient aspect comprising at least one of;

a job requirement not evidenced by the candidate ré

sumé

information, a job requirement not evidenced in candidate ré

sumé

information originally provided by the at least one candidate, or an ambiguity in the candidate ré

sumé

information; and

generating and sending, with the automated recruitment system, a recruiter recommendation message comprising a recommendation to fix the deficient aspect.

13. The method of claim 1, further comprising:

processing fourth instructions and information associated with a candidate rejection message received from the at least one candidate;

generating one or more preferred job requirements for the at least one candidate, the step of generating one or more preferred job requirements comprising at least one of;

analyzing information in the conversation and identifying preferred job requirements previously provided by the at least one candidate, or generating and sending a recruiter message requesting one or more job requirements to be provided by the at least one candidate and receiving a response message from the at least one candidate with the one or more preferred job requirements;

performing crawling, scraping, and parsing, with an automated recruitment system, of one or more job databases, and identifying a curated list of one or more alternative jobs having a job description having a matching score above a threshold value with the one or more preferred job requirements of the at least one candidate and with the candidate ré

sumé

information; and

generating and sending a message to the at least one candidate comprising the curated list.

14. The method of claim 13, further comprising:

identifying one or more finder'"'"'s-fee-eligible job databases in the one or more job databases, and generating an ordered list of the one or more finder'"'"'s-fee-eligible job databases by a size of a finder'"'"'s fee provided for a successful recruitment;

performing crawling, scraping, and parsing on a first finder'"'"'s-fee-eligible job database in the ordered list, and determining if the first finder'"'"'s-fee-eligible job database has at least one alternative job having a matching score above the threshold value; and

performing crawling, scraping, and parsing on another job database in the one or more job databases when the first finder'"'"'s-fee-eligible job database does not have at least one alternative job having a matching score above the threshold value.

15. An automated recruitment system, comprising a natural language processing engine and a database comprising a plurality of intents of different types and a plurality of entities of different types, the intents being natural language processing intents and the entities being natural language processing entities, the plurality of intents and the plurality of entities configured into a plurality of conversation paths likewise stored in the database, each of the plurality of conversation paths comprising an expression of at least one intent about at least one entity;

wherein the automated recruitment system is configured to;

identify at least one job description, at least one job requirement for the job description, and at least one candidate having candidate ré

sumé

information associated therewith;

initiate a conversation with the at least one candidate based on the plurality of conversation paths stored in the database;

process first instructions and information associated with a first intent that outputs a first message to the at least one candidate;

process second instructions and information associated with a candidate response message received from the at least one candidate, identify one or more candidate response intents in the candidate response message, and map the one or more candidate response intents to a conversation path in the plurality of conversation paths;

process third instructions associated with the conversation path in the plurality of conversation paths, and generate and provide to the at least one candidate a recruiter message based on the third instructions;

wherein the recruiter message comprises at least one of a question for the at least one candidate generated based on the candidate ré

sumé

information and a provision of further information to the at least one candidate based on an expression of candidate interest in the one or more candidate response intents.

16. The system of claim 15, wherein the system is configured to identify whether the one or more candidate response intents are casual intents, business intents, or other intents, the other intents being intents not classified as casual intents or business intents, and, when an intent in the one or more candidate response intents is identified as including at least one other intent, send a notification message to a recruiter regarding the other intent.

17. The system of claim 15, further comprising a plurality of communications systems including at least one text system and at least one speech system, and wherein the automated recruitment system is configured to send and receive candidate response messages using the at least one text system and the at least one speech system;

and

wherein the database is configured to store ongoing conversations with candidates in conversation-specific form such that candidate response messages received via the at least one text system and candidate response messages received via the at least one speech system are consolidated into a single ongoing conversation in the database.

18. The system of claim 15, wherein the automated recruitment system is configured to perform crawling, scraping, and parsing of searchable data, and is further configured to amend the candidate ré

sumé

information associated with the at least one candidate to incorporate at least one element of the searchable data.

19. The system of claim 15, wherein the automated recruitment system is further configured to:

analyze a level of similarity between the candidate ré

sumé

information and the at least one job requirement of the job description, and identify a deficient aspect of the candidate ré

sumé

information, the deficient aspect comprising at least one of;

a job requirement not evidenced by the candidate ré

sumé

information, a job requirement not evidenced in candidate ré

sumé

information originally provided by the at least one candidate, or an ambiguity in the candidate ré

sumé

information; and

generate and send a recruiter recommendation message to the at least one candidate comprising a recommendation to fix the deficient aspect.

20. The system of claim 15, wherein the system is further configured to:

process fourth instructions and information associated with a candidate rejection message received from the at least one candidate;

generate one or more preferred job requirements for the at least one candidate, the step of generating one or more preferred job requirements comprising at least one of;

analyzing information in the conversation and identifying preferred job requirements previously provided by the at least one candidate, or generating and sending a recruiter message requesting one or more job requirements to be provided by the at least one candidate and receiving a response message from the at least one candidate with the one or more preferred job requirements;

perform crawling, scraping, and parsing of one or more job databases;

identify a curated list of one or more alternative jobs having a job description having a matching score above a threshold value with the one or more preferred job requirements of the at least one candidate and with the candidate ré

sumé

information; and

generate and send a message to the at least one candidate comprising the curated list.