Crafted Works, Bold Results

Developed a fully automated end-to-end image pipeline in Python: built high-performance web scrapers to collect thousands of graphic assets, implemented AI-powered spell-checking and quality validation using OpenAI GPT-4 Vision, performed content regeneration and metadata enrichment with LLMs (GPT-4o-mini), stored and managed data in MongoDB Atlas, and exposed the curated dataset via a production-ready Flask API. The system includes robust error handling, logging, rate-limiting, and filename normalization across multiple processing stages.

Talent Intelligence Automation

Designed and deployed a fully automated LinkedIn talent intelligence pipeline using n8n: scraped 10,000+ public profiles, extracted and structured 30+ fields (skills, experience, education, location, endorsements, job titles, company history, summary, recommendations, etc.), normalized messy data with regex + OpenAI function calling, and exported perfectly categorized CSVs/JSON for recruiting and lead-gen databases.

AI Voice Hiring System

Built an end-to-end voice AI recruiting system: candidate sourcing → personalized cold-call scheduling via SMS/email → live phone interviews with natural speech → real-time sentiment & qualification scoring → automated follow-up + summary report sent to hiring managers.