AI Resume Search Chatbot

I need a working prototype that turns a folder of PDF resumes into a searchable talent database and then lets a natural-language chatbot surface the best-fit profiles. Everything must be built in Python and rely on solid AI/ML techniques.

Scope
• Parse the combination-style (chronological + functional) PDFs, clean the text and store it in a query-ready database or vector store.
• Enable the chatbot to understand a recruiter’s free-form questions, extract the right intent (skills, years of experience, location, etc.), pull structured data from the resumes, and rank candidates by relevance.
• The NLP layer must therefore handle three tasks fluently: understanding user queries, extracting specific information from resumes, and ranking & matching candidates.

What I expect to receive
1. Well-commented Python code (resume ingestion, embedding/indexing, and conversational retrieval pipeline).
2. A small, sample resume database so I can run end-to-end tests.
3. A concise document that walks through every design decision—including why each algorithm, model, or library was chosen—and the exact steps to reproduce the environment.
4. A short screencast or demo video that shows the chatbot answering a few realistic recruiter questions and displaying ranked results.

Acceptance criteria
• The chatbot returns relevant candidates in a single turn for queries like “Show me Python developers with 5+ years in finance.”
• Ranking clearly reflects the stated intent (skills and experience weighting is obvious).
• I can recreate the database and rerun the demo by following your documentation with no missing steps.

Mention any preferred libraries or frameworks in your proposal—I’m open as long as everything stays in Python and the reasoning is transparent.