Automate Job applications 🧑🏻‍💼📄with Chatbot, PredictLeads and ScrapegraphAI

This Chatbot automates the process of discovering job openings and generating tailored job application emails.

It combines AI agents, web scraping, and email drafting to streamline job applications.

This workflow transforms job applications from a manual, repetitive process into an intelligent AI-assisted automation system that:

Saves time Improves email quality Reduces errors Maintains human oversight Scales across multiple job postings

It represents a strong example of combining conversational AI, external data tools, structured parsing, and workflow automation into a production-ready solution.

How it works

User starts a chat – The workflow begins when a user sends a message via the chat trigger.

PredictLeads Agent processes the request – A LangChain agent determines the user's intent. If the request involves company research, it first queries Context7, then optionally PredictLeads for deeper data.

Response parser – The agent's output is cleaned and parsed into a structured JSON format with list (boolean) and output fields.

List check – If list is true (e.g., a list of job URLs), the workflow extracts links and passes them to the next stage. If false, the agent responds directly to the user.

Link extraction – The Links Extractor node uses OpenAI to extract job posting URLs from the user's input.

Loop through each link – Each URL is processed individually using a Loop Over Items node.

Scrape job details – The Scrape Job node (powered by ScrapegraphAI) extracts: Email address to send the application to Job position title Full job description text

Email presence check – If an email is found, the workflow proceeds to generate an application email. If not, it informs the user that no email is available and provides the job link.

Job Application Agent – A Gemini-powered agent generates a professional email using: Candidate's personal info (name, location, skills) Job position and description A tool (Create email) to format the subject and body

Send email tool – The agent triggers the Send email workflow, which: Fetches the CV from a public URL Creates a draft in Gmail with the CV attached

User response – The final output is sent back to the user via chat, confirming the draft creation or notifying them of missing information.

Setup steps

To use this workflow, you need to configure the following credentials and nodes:

1. Chat Trigger No setup required. This is the entry point for user messages.

2. OpenAI Chat Model Add your OpenAI API key.

3. Google Gemini Chat Model Add your Google AI API key.

4. Context7 MCP Tool Credential**: Context7 Add your API key as a header (e.g., Authorization: Bearer XXX).

5. PredictLeads MCP Tool Credential**: Multiple Headers PredictLeads Add required headers (e.g., X-API-Key or similar).

6. ScrapegraphAI Add your ScrapegraphAI API key.

7. Gmail Authorize access to Gmail (OAuth2) to create drafts.

8. HTTP Request (Get CV) Ensure the CV is publicly accessible at the URL in the node (https://XXX/cv.pdf) or update it with your own.

9. Simple Memory No setup needed. Used to maintain conversation context.

10. Agent Prompt Customization (Optional) Review the system prompts in the PredictLeads Agent and Job application Agent nodes. Update candidate personal information (name, location, etc.) in the Job application Agent prompt.

11. Workflow ID for "Send email" The Send email tool calls another workflow by ID . Ensure this ID matches the current workflow (it should be self-referential).

Key Advantages

1. ✅ End-to-End Automation

It automates the entire job application lifecycle:

Job discovery Job data extraction Email writing CV attachment Draft preparation

No manual copy-paste required.

2. ✅ AI-Orchestrated Tool Usage

The system intelligently decides when to use:

Company research tools (Context7) PredictLeads data Scraping services Email drafting workflows

This makes it dynamic and adaptable rather than static.

3. ✅ Structured & Reliable Data Handling

Uses JSON schema validation Cleans malformed AI outputs Ensures consistent structured results Reduces errors in automation flows

4. ✅ Human-in-the-Loop Safety

Before sending any email:

The system requires double approval The email is saved as a draft, not automatically sent

This prevents accidental or incorrect applications.

5. ✅ Personalized & Tailored Applications

Each application is:

Context-aware Position-specific Professionally formatted Generated using candidate-specific data

This increases response quality compared to generic templates.

6. ✅ Scalability

Because of:

Split-in-batches logic Looping over multiple job listings Structured parsing

The workflow can process multiple job opportunities efficiently.

7. ✅ Modular Architecture

The workflow is cleanly modular:

AI agents Scraper Parser Email tool CV fetcher

👉 Subscribe to my new YouTube channel. Here I’ll share videos and Shorts with practical tutorials and FREE templates for n8n.

Need help customizing?
Contact me for consulting and support or add me on Linkedin.