This project automates the process of generating cold emails tailored to specific job roles by extracting relevant skills and experience from a portfolio and matching them to job descriptions. It leverages the power of language models and similarity searches for precise and contextually aware email generation.
- Web Scraping: Scrapes relevant textual data from job descriptions.
- Tech Stack Matching: Uses a vector database to match skills in the job description to the candidate’s experience and past projects.
- AI-Powered Prompts: Meta AI’s Llama 3.1 LLM generates personalized emails based on the role, skills, and experience.
- Streamlit Frontend: User-friendly interface that allows users to input job links and see generated emails.
- Chroma for Similarity Search: Performs similarity searches to find the best match between the job description and the candidate’s skillset.
- Input: Users provide a job link through the Streamlit frontend.
- Web Scraping: The link is processed to extract textual job data.
- Tech Stack Extraction: The extracted job data is processed to identify required skills and experience, which are stored in a CSV file.
- Vector Database: The portfolio's tech stack and job skills are stored in Chroma for a similarity search.
- AI Processing: A JSON file containing the role, skills, and description is fed into Meta AI’s Llama 3.1 LLM, generating a personalized cold email prompt.
- Output: The generated email is displayed on the frontend.
- Frontend:
- Streamlit
- Backend:
- Meta AI’s Llama 3.1 LLM
- ChromaDB for vector database and similarity search
- Python/Langchain/BeatifulSoup for web scraping and data processing