We noticed that recruiters and sales teams were wasting tons of time sifting through LinkedIn profiles, so we decided to build two agents that could automate the process.
LinkedIn Lead Agent
We built this agent to search for LinkedIn profiles using Tavily and then parse the profiles to pull out key info like names, locations, experience, education, and skills. The fun part was writing our parse_linkedin_profiles function that digs into raw profile data and neatly formats everything into a clear, numbered list.
Profile Analyzer Agent
The second agent takes the profiles collected by the first one and looks for trends. It aggregates data to spot common skills and industries, giving users a quick snapshot of potential leads and networking opportunities.
Working on these agents during the hackathon was all about rapid prototyping and creative problem-solving. We faced a few challenges with inconsistent profile formats, but each obstacle made our solution even better. In the end, our two agents teamed up to turn messy data into actionable insights.
Built With
- fetch
- langchain
Log in or sign up for Devpost to join the conversation.