Scrapegraph AI-powered Job Application assistant

Inspiration

Job applicants spend hours manually searching for company contacts and preparing for interviews, often jumping across LinkedIn, blogs, and interview forums. I wanted to build an AI assistant that automatically gathers this information and turns it into a clean preparation report.

What it does

The tool takes a company and job role as input and automatically recommends relevant jobs & finds relevant employees to reach out to and collects interview questions from the web. It then summarizes everything into a clear, human-readable preparation guide for the candidate.

How we built it

The system uses Tavily search to discover relevant pages, ScrapeGraphAI APIs to extract structured data, and an LLM pipeline orchestrated with LangGraph to analyze, rank contacts, and generate a final interview preparation summary. ScrapeGraphAI enables AI-driven extraction from websites without needing manual HTML selectors. The recommended jobs were ingested in Supabase & profile was analyzed.

Challenges we ran into

One challenge was reliably extracting structured information from different websites with varying layouts. Another was designing a multi-step pipeline that coordinates search, scraping, and summarization while maintaining traceable source URLs.

Accomplishments that we're proud of

We built a fully automated job research pipeline that can gather contacts and interview preparation material in seconds. The system also returns source links for transparency, making the output useful and trustworthy for real job applicants.

What we learned

We learned how powerful LLM-driven scraping and agent pipelines can be when combined with structured extraction tools like ScrapeGraphAI. Designing modular subgraphs also makes it easy to plug this capability into larger AI workflows.