What does the recruiting AI agent do?

It handles the research-heavy parts of hiring: screening resumes against a job description, ranking candidates with reasons, analyzing and summarizing LinkedIn profiles, scanning the hiring market for a role, sourcing passive candidates from open-source projects and public work, and reporting time in stage across your pipeline. Results are written back to your tracker and summarized in chat.

Can it screen resumes against a job description?

Yes. Paste the job description and share the resumes as a sheet, an ATS export, uploaded files, or profile URLs. The agent scores every candidate against the must-haves and nice-to-haves, cites the evidence for each score, flags claims it could not verify, and recommends advance, discuss, or decline. You can change the weighting or add a hard requirement at any point.

Can it analyze or summarize a LinkedIn profile?

Yes. Give it a profile URL and it scores the experience against the role and writes a one-page brief covering trajectory, scope of ownership, notable work, and gaps. Teams use the same brief for interview panels.

Does it connect to my ATS?

It works from any export your ATS produces and writes results to Google Sheets out of the box. Connectors for tools like Ashby and Greenhouse can be added in the agent settings, and the agent reads and writes through them once connected.

Will it make hiring decisions or reject candidates on its own?

No. It recommends and explains. People decide. Nothing is rejected automatically, every score comes with its reasoning, and the agent evaluates only against the criteria you provide. That keeps the process consistent and auditable, and it is how we recommend you run it.

Can it track how long candidates spend in each stage?

Yes. Share a pipeline export and it calculates days per stage per role, names the bottleneck stage, highlights stuck roles, and can post the summary to a Slack channel every week.

Recruiting AI Agent

A recruiting AI agent that screens, scores, and sources candidates

Name: Recruiting AI Agent
Author: Gumloop

Share a job description and your candidates. The agent screens every resume, scores LinkedIn profiles, ranks the shortlist with evidence, and writes it back to your tracker.

Book a demo

Works with

Explore 100+ integrations

Trusted by leading teams

Case study

Make it yours

Start with a working agent, then shape it to your team

The Recruiting AI Agent comes set up with these apps, instructions, and prompts. Change any of them once it's in your workspace.

Apps it connects to

Google Sheets

Instructions it follows

Score only against the stated requirementsUse the must-haves and nice-to-haves provided. Never infer requirements or weigh anything unrelated to the role.
Evidence or flagEvery score cites the resume line or public source behind it. Claims that cannot be verified are flagged for a human.
Recommend, do not decideOutput advance, discuss, or decline with a reason the recruiter can reuse. Never mark a candidate rejected or send anything unapproved.

Try asking

Screen the candidates in this sheet against the job description in the first tab and rank them.
Analyze this LinkedIn profile for a Head of Marketing role and write a one-page panel brief.
Who is hiring senior data engineers in Toronto right now, what do they require, and what do they pay?
Find the top 20 contributors to these three open-source repos and evaluate hireability.

The first screen in minutes, not days

Screening, profile analysis, sourcing, and pipeline reporting, scored with evidence, written to your tracker.

Without Gumloop

With Gumloop

Without vs. With Gumloop

A good role draws 300 applications and the first screen takes a recruiter two days.

The agent reads every resume against the requirements and hands back a ranked shortlist with the evidence, in minutes.

Candidate research is 20 browser tabs per profile and notes nobody can find later.

It pulls LinkedIn, GitHub, portfolio, and talks into one summary per candidate and writes it to your tracker.

Sourcing passive candidates is the highest-leverage work and the first thing that gets dropped.

Ask for open-source contributors or speakers in a niche and get a sourced list with a reason for each and a draft first message.

Hiring managers ask how long candidates sit in each stage and the answer is always "let me pull that".

Point it at the pipeline export and get days per stage, the bottleneck named, posted to Slack weekly.

What is the Recruiting AI Agent?

The Gumloop recruiting AI agent takes the repetitive parts of hiring off a recruiter: resume screening against role requirements, candidate ranking, LinkedIn profile analysis, market research for a role, passive candidate sourcing, and pipeline reporting. Give it a job description and a set of candidates, as a spreadsheet, an ATS export, uploaded resumes, or profile URLs, and it scores every candidate and shows the evidence.

It does the research a recruiter would do before a screen. It reads the resume, checks the public record on LinkedIn, GitHub, and the open web, verifies the claims it can, flags the ones it cannot, and writes one summary per candidate. Results land back in the tracker you already use, so the hiring manager sees a ranked shortlist with reasons instead of a folder of PDFs.

The agent recommends and people decide. It never rejects a candidate on its own, it explains every score in plain language, and it scores only against the requirements you give it. Run it on one role, point it at the whole pipeline for a time-in-stage report, or ask it to source candidates from open-source projects and conference lineups when inbound is thin.

What you can do with the Recruiting AI Agent

Workflows the agent handles out of the box.

Resume screening against the role

Reads every resume against the must-haves and nice-to-haves and returns a score, the evidence for it, and a recommended next step.

Candidate ranking with reasons

Sorts the pool into advance, discuss, and decline with one line each, so recruiters can reply honestly and hiring managers can skim.

LinkedIn profile analysis

Scores a profile against the role and writes a one-page brief: trajectory, scope, notable work, and the questions to ask.

Hiring market scanner

Researches who else is hiring for the role, what they require, and what they pay, so the job description and the offer match the market.

Passive candidate sourcing

Finds open-source contributors, speakers, and builders in a niche, evaluates fit from public work, and drafts a first message.

Time-in-stage reporting

Reads a pipeline export and reports days per stage, where candidates drop, and which roles are stuck, posted to Slack on a schedule.

How to use the Recruiting AI Agent

Get from landing page to live agent in a few clicks.

1
Click "Use this agent"
An agent is created in your Gumloop workspace with the screening rules loaded, web search on, and Google Sheets ready to connect for your tracker.
2
Share the role and the candidates
Paste the job description and link a sheet, upload resumes, or paste profile URLs. Say which requirements are hard.
3
Get the shortlist
Scores, reasons, and next steps are written to your tracker and summarized in chat. Ask for invites, decline notes, or a panel brief next.

Recruiting AI Agent use cases

Real workflows teams run with this agent.

Screen resumes against role requirements

Job description plus a folder or sheet of resumes in, every candidate scored against the must-haves out, with evidence.

Rank candidates for the hiring manager

Turn 300 applicants into advance, discuss, and decline with one honest line per candidate.

Analyze and score a LinkedIn profile

Paste a profile URL and get the experience scored against the role plus a summary of trajectory, scope, and gaps.

Write a candidate brief for the panel

Before an onsite, one page per candidate from their profile and public work, so every interviewer walks in prepared.

Scan the hiring market for a role

See who else is hiring for it, what they require, and what they pay before you publish the job description.

Source open-source contributors

Find the most active contributors to the libraries your team depends on and get a first message drafted for each.

Report time in stage across the pipeline

Point it at an ATS export and get days per stage, the bottleneck per role, and a Slack summary every week.

Prep interviewers and candidates

Draft a structured scorecard for the panel, interview prep for a candidate, and outreach that reads like a person wrote it.

Why use Gumloop for the Recruiting AI Agent

Recommends, never decides

Every score is explained and nothing is rejected automatically. The decision and the audit trail stay with your team.

Scores against your criteria only

It evaluates against the requirements you give it, not a hidden idea of a good candidate.

Lands in the tracker you use

Results write back to the sheet or doc your team works from. ATS connectors are available.

Change the rules in chat

Add a hard requirement, change the weighting, switch roles, or ask for outreach in your voice.