DimiMikadze/orca

What it solves

Orca solves the problem of manually analyzing LinkedIn profiles to find deep, qualitative insights. Instead of skimming a profile, it automates the extraction of structured data such as a person's pain points, values, expertise, and communication style based on their actual activity.

How it works

Orca takes a LinkedIn profile URL and a set of desired insights. It scrapes baseline data including posts, comments, and reactions. An AI agent then reasons over this data to extract the requested insights; if the agent determines it needs more information to reach a conclusion, it autonomously calls additional scraping tools to gather more data.

Who it’s for

• Sales professionals looking to understand a prospect's priorities before outreach.
• Recruiters wanting to assess a candidate's genuine interests beyond their resume.
• Investors seeking to map a founder's thinking and evaluate their positioning.

Highlights

• Autonomous data gathering: The agent can decide to call more scraping tools if the initial data is insufficient.
• Structured insights: Extracts specific attributes like network influence and how interests evolve over time.
• Integration ready: The core logic is available as a standalone Node.js library.
• LLM flexibility: Built with LangChain to support multiple providers including OpenAI and Anthropic.