Automate Your Recruiting

Transform manual tasks that take hours into instant workflows powered by InternGlobe's Artificial Intelligence framework.

Why automate?

Recruiting college students or recent graduates often attracts massive volumes of applications for a single job posting. Taking the time to open each resume, read it, evaluate it, and then write a custom email (rejection or invitation) by hand can consume half of a recruiter's entire week. With InternGlobe Workflows, you are not displacing humans: you are freeing them up to completely focus on high-value face-to-face interviews.

Massive Time Savings

An application is read, scored at 64/100, placed in "Average", and notified by a very cordial email pointing out improvement areas... all in exactly 4.2 seconds. Without a single click from you.

Unbiased AI Decisions

No more punishing "diagonal reading" because you are short of time. Globy reads the entire profile objectively by applying the "Company Culture" you instructed.

5-Star Candidate Experience

No more applicants left "ghosted" for weeks. They get high-quality, insightful replies rapidly, and your employer branding improves exponentially.

System Anatomy of a Workflow

You build a Workflow visually on a "Canvas". It's simply a flow-chart where small boxes (called nodes) are linked together using cables (called edges).

1. The Trigger (Anchor point)

The root spark. Every flow begins with an indisputable event. For example: "A student drops their resume on the Data Scientist posting" or "14 days of inactivity on a case".

2. The Intelligence (Globy Orchestrator)

The thinking phase (optional but powerful). Globy intercepts the Trigger profile and employs sub-agents to: condense an endless PDF, generate a text, evaluate a trait, or translate an asset.

3. Logic (Condition Crossroads)

The switch tracks of the flow. If Globy determines the profile's score > 80, the flow rides the True edge (path A). Otherwise, it hits False (path B). Logic also involves timers (Pause 3 days to follow up).

4. Final Action

The physical outcome that a human intern would have executed themselves: delivering the end-to-end email (statically written or natively by AI), attaching a UI label, or progressing the stage (Review ➜ Terminated).

Engineering Catalogue

Now that you understand the architectural philosophy backing recruiting automation, below is the granular data-dictionary of all Nodes and Variables you wield to form your pipelines.

Dynamic Variables

Use these tags in your emails and notifications to personalize your messages automatically:

{'{{'}candidateName{'}}'}

Prénom et nom du candidat

{'{{'}internshipTitle{'}}'}

Titre du stage concerné

{'{{'}employerName{'}}'}

Nom de votre entreprise

{'{{'}globyScore{'}}'}

Score de matching généré par l'IA

{'{{'}applicationId{'}}'}

ID unique de la candidature

1. Triggers

New Application

Fires immediately when a resume is submitted.

Configuration

  • Target Listing: All listings or a specific one by ID.
  • Listing ID: Technical ID of the job post (found in URL).

Output Data

Output Variables: {{applicationId}}, {{candidateEmail}}.

Stage Changed

Triggers when a candidate moves across your pipeline phases.

Configuration

  • From Stage: The initial stage (optional).
  • To Stage: The stage that fires the workflow (e.g., Interview).

Candidate Waiting

Fires when a profile has been waiting for a response for too long.

Configuration

  • Waiting Days: Number of consecutive days with no action on the profile.

2. Actions

Email to Candidate

Automated delivery of a customized email.

Subject

Email subject line (supports variables).

Message

Rich content with dynamic values.

Reply To

Custom return email address.

Update Stage

Auto-move candidate across CRM stages.

New Stage

The target phase (e.g., Rejected, Hired).

Confirm

Send native platform confirmation to candidate.

Add a Tag

Marks the candidate profile with a visual tag.

Tag Name

The text of the tag (e.g., Talent Pool, Top Profile).

3. Logic & Conditions

IF / ELSE Logic

Logic fork for your workflow.

Test Parameters

Field

Property to test (Stage, AI Score, Resume...).

Operator

Comparison (Equals, Greater than, Contains...).

Value

The target for comparison.

Expected Behavior

If condition is met -> "True" Path
Else -> "False" Path

4. The Globy AI Orchestrator & Sub-Agents

Globy is not just a simple node. It is an Intelligent Orchestrator that delegates work to specialized sub-agents. The orchestrator enforces your Company Culture and Strict Rules to every agent down the line.

Globy Orchestrator

globy_orchestrator

The central brain directing AI sub-agents. Defines the tone, strict rules, and company culture to respect for the entire flow.

System PromptThe main role of the Agent (e.g. You are an HR Expert).
Strict RulesWhat the orchestrator must absolutely NEVER do.
Company CultureThe tone to adopt (dynamic, formal, friendly).
TasksThe sequence of sub-tasks to delegate to sub-agents.

Sub-agent: Scoring

globy_score

Scores the application from 0 to 100, justifying its choice based on your strict criteria.

  • CriteriaKey points to scrutinize (e.g. +20 points if React JS).
  • Min ThresholdThe minimum score required to move to the next step.

Sub-agent: Drafting

globy_draft

Drafts perfect emails (rejection, interview invite) adapting to the candidate's journey.

  • PromptThe goal of the email (e.g. Draft an encouraging rejection).
  • ToneWarm, Direct, Formal.
  • TargetCandidate (external) or Employer (internal).

Sub-agent: Summary

globy_summarize

Condenses a 3-page resume into an ultra-fast summary for the recruiter to read.

  • FormatParagraph, Bullet points, or Scorecard.
  • FocusWhat to look for (Strengths, Weaknesses).

Sub-agent: Classify

globy_classify

Automatically sorts complex profiles into your own categories.

  • CategoriesComma-separated list (e.g. Tech, Marketing, Data).
  • Sorting RulesHow the AI should make its final decision.

Sub-agent: Translation

globy_translate

Translates any document or generated summary.

  • FromSource language (Auto-detect supported).
  • ToTarget language (French or English).

Simulator & Debugging

// Simulating a score of 85 application

09:00:01 [START] Trigger: New Application (John Doe)

09:00:02 [NODE] AI_SCORE -> Result: 85/100

09:00:03 [LOGIC] IF (score > 80) -> Output: TRUE

09:00:04 [ACTION] Email Candidate (Success)

09:00:05 [FINISH] Workflow complete in 4.2s

Need specific guide?

Our technical documentation evolves weekly with new nodes.

Recruitment Automations - InternGlobe | InternGlobe