:Finding the right job or candidate manually takes hundreds of hours. To solve this, I engineered a fully autonomous job-hunting engine using n8n, custom API integrations, and Large Language Models (LLMs).

This workflow acts as a 24/7 AI Recruiter. It continuously pulls live job postings from 4 different data sources, filters out the noise, and uses an AI Agent to read the job descriptions and compare them against a specific CV profile.

TECHNICAL CHALLENGES SOLVED

• Multi-Source Aggregation: Integrated 4 separate HTTP data streams and normalized the messy JSON data into a single, clean pipeline using custom JavaScript nodes. • AI Decision Making: Connected OpenRouter & Ollama LLMs directly into the flow. The AI acts as the "Brain," reading complex job requirements and outputting structured JSON decisions (Match / No Match). • Resilient Error Handling: Built robust error-catching logic that sends immediate Telegram alerts if any API endpoint fails, ensuring the system never fails silently. • High-Volume Processing: Utilized Batch/Loop nodes to process hundreds of job postings per minute without hitting API rate limits.

RESULTS & BUSINESS VALUE

✓ Replaced 20+ hours a week of manual job searching and filtering. ✓ Achieved near-instant notification of new, highly-relevant job postings. ✓ Built a scalable architecture that can easily be adapted for real estate, lead generation, or news aggregation.

TECH STACK

n8n • JavaScript • REST APIs • LLMs (OpenRouter/Ollama) • Telegram Bot API • Webhooks • Prompt Engineering