Recruiting · Production

Under the hood: the tech behind LinkedIn Hiring Assistant, LinkedIn's first AI recruiting agent

The problem

Recruiters had to navigate multiple tools and handle many repetitive tasks—sourcing, messaging candidates, coordinating with hiring managers—leaving limited time for strategic and interpersonal work.

Workflow diagram · grounded in source
1
Recruiter natural language input
trigger
“working with natural language as an input”
2
AI job description drafting
ai_action
“Building a job description and refining it via collaborative dialogue with the recruiter”
3
Requirements to search query translation
ai_action
“Translating and deeply understanding explicit and implicit requirements into recruiter search queries”
4
Candidate ranking and matching
ai_action
“Refining and ranking search results by cross-referencing preferred and required qualifications against candidates' profiles and resumes, including explanations of how and where they match”
5
Recruiter feedback and review
human_review
“Allowing the recruiter to provide feedback to iteratively refine the list of potential candidates”
6
Experiential memory adaptation
feedback_loop
“experiential memory, which is its ability to learn from its activity or interactions with each individual recruiter. This means that over time, Hiring Assistant will also become more personalized to each recruiter, learning from their fe…”
7
Candidate pipeline output
output
“Helping translate recruiter notes into a pipeline of qualified candidates or manage interview coordination”
Reported outcome

LinkedIn launched Hiring Assistant, an AI agent that takes on recruiters' most repetitive tasks using LLMs, experiential memory, and semantic search, enabling recruiters to focus on strategic and interpersonal work.

Reported metrics
Recruiter task focus shiftfocus on more strategic, interpersonal aspects of work
Recruiter repetitive task loadtaking on their most repetitive tasks
Reported stack
LinkedIn Hiring AssistantLLMsLinkedIn Economic Graph
Source
https://www.linkedin.com/blog/engineering/generative-ai/the-tech-behind-the-first-agent-from-linkedin-hiring-assistant
Read source ↗

Frequently asked questions

What did this team achieve with this AI workflow?

LinkedIn launched Hiring Assistant, an AI agent that takes on recruiters' most repetitive tasks using LLMs, experiential memory, and semantic search, enabling recruiters to focus on strategic and interpersonal work.

What tools did this team use?

LinkedIn Hiring Assistant, LLMs, LinkedIn Economic Graph.

What results were reported?

Recruiter task focus shift: focus on more strategic, interpersonal aspects of work; Recruiter repetitive task load: taking on their most repetitive tasks (source-reported, not independently verified).

How is this recruiting AI workflow structured?

Recruiter natural language input → AI job description drafting → Requirements to search query translation → Candidate ranking and matching → Recruiter feedback and review → Experiential memory adaptation → Candidate pipeline output.

Related recruiting cases