| import faiss | |
| import numpy as np | |
| from embeddings import generate_embeddings | |
| dimension = 384 | |
| index = faiss.IndexFlatL2(dimension) | |
| resume_db = {} | |
| def store_resume_embedding(filename, vector, text): | |
| global resume_db | |
| index.add(np.array([vector])) | |
| resume_db[len(resume_db)] = {"filename": filename, "text": text} | |
| def search_similar_resumes(job_description): | |
| job_vector = generate_embeddings(job_description) | |
| D, I = index.search(np.array([job_vector]), k=3) | |
| return [resume_db[i]["text"] for i in I[0] if i in resume_db] | |