United States
Web (n8n Automation)
AI-Powered Workflow Automation
n8n + GPT-4o + PostgreSQL
Overview
AI-Powered Job Discovery & Company Research, Fully Automated
Finding the right remote Flutter and Full Stack Developer roles across the US and Europe is a time-consuming, repetitive grind. You're checking LinkedIn, scanning job boards, filtering out irrelevant listings, and then manually researching each company before deciding whether to apply.
We built a fully automated n8n workflow system that runs 24/7 — discovering jobs from LinkedIn and multiple web sources, filtering them by geography, enriching each listing with AI-extracted salary and requirements data, and then deploying an AI research bot to generate scored intelligence reports on every hiring company.
The system replaces hours of daily manual job hunting with a hands-off pipeline that delivers pre-researched, pre-scored job opportunities directly to Slack.
Tech Stack
The Problem
What Manual Job Hunting Looks Like
Manually searching LinkedIn, RemoteOK, We Work Remotely, and Google Jobs multiple times per day
Sifting through hundreds of listings from regions outside the target market
No structured way to extract salary ranges or candidate expectations from unstructured descriptions
Zero visibility into the company behind the job - size, funding, tech stack, culture, or red flags
Duplicate listings across platforms wasting valuable time
Missed opportunities because high-quality listings expire before they're seen
Architecture
3 Interconnected n8n Workflows
The system is split into three modular workflows for reliability and clear separation of concerns. Each workflow can be monitored, debugged, and scaled independently.
Job Fetching & Filtering
Trigger: Cron schedule - every 6 hours
- LinkedIn Jobs API (OAuth2) searches for Flutter & Full Stack Developer positions
- Parallel HTTP requests to RemoteOK, Google Jobs, We Work Remotely
- Data normalization into unified schema across all sources
- Region filter engine with country/city blocklist
- Deduplication against PostgreSQL using composite keys
AI Enrichment & Storage
Trigger: Called by Workflow 1 via Execute Workflow
- GPT-4o-mini extracts salary range, currency, and frequency
- GPT-4o summarizes candidate expectations and required skills
- GPT-4o detects hiring urgency and timeline signals
- Enriched records upserted to PostgreSQL jobs table
- New jobs trigger Workflow 3 with company details
AI Company Research Bot
Trigger: Called by Workflow 2 for new companies
- LinkedIn Company API fetches size, industry, HQ, founding date
- AI Agent researches funding, Glassdoor reviews, tech stack, culture
- GPT-4o generates structured intelligence report with 0-10 score
- Report stored in company_research table linked to all job listings
- Score 7.0+ triggers formatted Slack notification
What We Built
8 Core Capabilities
Multi-Source Job Fetching
Parallel HTTP requests to LinkedIn, RemoteOK, Google Jobs (SerpAPI), and We Work Remotely. All results normalized into a unified schema regardless of source format.
Intelligent Region Filtering
Geographic blocklist engine filters jobs by location and company origin. Only US, Canada, and Western/Northern European positions pass through.
Cross-Source Deduplication
Composite key deduplication (source + external_id) against PostgreSQL ensures no duplicate listings across any of the four data sources.
AI Salary Extraction
GPT-4o-mini parses unstructured job descriptions to extract salary ranges, currency, and frequency. Handles "$120K-$160K", "competitive salary", and buried compensation details.
AI Company Research Bot
Autonomous AI agent conducts web research on every new company: funding, Glassdoor ratings, tech stack, culture signals, and red flags. Generates scored intelligence reports.
Slack Alerts for Top Matches
Companies scoring 7.0+ automatically trigger formatted Slack notifications with job title, salary range, company score, and direct application link.
AI Urgency Detection
Third AI extraction pass identifies hiring timeline signals: "start ASAP", specific dates, or standard pipeline timelines. Categorized as Immediate, 2 Weeks, 1 Month, or Not Specified.
Error Handling & Retry Logic
Connected to n8n Error Handler workflow. Rate limit errors trigger exponential backoff. API failures are logged and the pipeline continues processing remaining sources.
Sample Output
AI-Generated Company Report
| Industry | FinTech |
| Size | 51-200 employees |
| HQ | San Francisco, CA |
| Funding | Series B - $45M |
| Glassdoor | 4.2 / 5.0 |
| Culture | Remote-first, async-heavy, engineering-led |
| Red Flags | High engineering turnover in Q3 2025 |
| Overall Score | 8.2 / 10 |
| AI Recommendation | Strong match. Growing company with active Flutter investment... |
Data Layer
PostgreSQL Schema
jobs
Unique: (source, external_id) | Indexed: source, company_name, status
company_research
Unique: (company_name) | Indexed: company_name, overall_score
Outcome
What Gets Delivered
Fully automated job discovery pipeline running every 6 hours, 24/7
Multi-source coverage: LinkedIn + 3 web job boards aggregated
AI-enriched records with structured salary, requirements, and urgency
Geographic filtering eliminates irrelevant regions automatically
Cross-source deduplication - zero repeated listings
AI-generated company intelligence reports with quality scores
Slack alerts for high-quality matches (company score 7+)
Modular 3-workflow architecture - independently deployable
Need an AI-powered automation workflow built for your business?
