Update app.py

app.py

@@ -6,7 +6,9 @@ from PyPDF2 import PdfReader
 import docx
 import re
 import google.generativeai as genai
+import time
 import concurrent.futures
+from fuzzywuzzy import fuzz
 
 # Load pre-trained embedding model for basic analysis
 sentence_model = SentenceTransformer('paraphrase-MiniLM-L6-v2')
@@ -82,27 +84,40 @@ def extract_text_from_file(file_path):
         return ""
 
 def analyze_with_gemini(resume_text, job_desc):
-    # Modified prompt to have Gemini calculate match percentage
     prompt = f"""
-    Analyze the following resume and calculate the overall match percentage for the job description. 
+    Analyze the resume with respect to the job description.
     Resume: {resume_text}
     Job Description: {job_desc}
-    
-    Provide:
+    Extract:
     1. Candidate Name
     2. Email Address
     3. Contact Number
     4. Relevant Skills
     5. Educational Background
-    6. Leadership Experience (years)
+    6. Team Leadership Experience (years)
     7. Management Experience (years)
-    8. Calculated Overall Match Percentage
-    
-    The Overall Match Percentage should represent how well the resume matches the job description, considering leadership, management, and skills.
+    8. Management Skills (e.g. strategic planning, team management, project management, etc.)
+    9. Match Percentage (leadership and management focus)
+    Provide a summary of qualifications in 5 bullet points.
     """
     response = genai.GenerativeModel('gemini-1.5-flash').generate_content(prompt)
     return response.text.strip()
 
+def extract_management_details(gemini_response):
+    leadership_exp_pattern = r"Team Leadership Experience \(years\):\s*(\d+)"
+    management_exp_pattern = r"Management Experience \(years\):\s*(\d+)"
+    management_skills_pattern = r"Management Skills\s*[:\-]?\s*(.*?)(?=\n|$)"
+
+    leadership_match = re.search(leadership_exp_pattern, gemini_response)
+    management_match = re.search(management_exp_pattern, gemini_response)
+    skills_match = re.search(management_skills_pattern, gemini_response)
+
+    leadership_years = int(leadership_match.group(1)) if leadership_match else 0
+    management_years = int(management_match.group(1)) if management_match else 0
+    skills = skills_match.group(1) if skills_match else ""
+
+    return leadership_years, management_years, skills
+
 def extract_candidate_details(gemini_response):
     name_pattern = r"Candidate Name\s*[:\-]?\s*(.*?)(?=\n|$)"
     email_pattern = r"Email Address\s*[:\-]?\s*(.*?)(?=\n|$)"
@@ -118,6 +133,46 @@ def extract_candidate_details(gemini_response):
 
     return name, email, contact
 
+def calculate_role_score(role_keywords):
+    seniority_score = 0
+    role_hierarchy = {
+        "CEO": 5,
+        "CIO": 5,
+        "Director": 4,
+        "VP": 4,
+        "Manager": 3,
+        "Team Lead": 2,
+        "Junior": 1
+    }
+
+    for keyword, score in role_hierarchy.items():
+        if fuzz.partial_ratio(keyword.lower(), role_keywords.lower()) > 80:
+            seniority_score = max(seniority_score, score)
+    
+    return seniority_score
+
+def calculate_advanced_match(leadership_years, management_years, skills, required_skills, role_keywords, max_leadership_exp=10, max_management_exp=10):
+    leadership_weight = 0.35
+    management_weight = 0.35
+    skills_weight = 0.2
+    role_weight = 0.1
+
+    leadership_score = min(leadership_years / max_leadership_exp, 1.0) * 100
+    management_score = min(management_years / max_management_exp, 1.0) * 100
+
+    role_score = calculate_role_score(role_keywords)
+    role_score = role_score * 100
+
+    skills_matched = sum(1 for skill in required_skills if fuzz.partial_ratio(skill.lower(), skills.lower()) > 80)
+    total_skills = len(required_skills)
+    skill_match_score = (skills_matched / total_skills) * 100
+
+    overall_match = (leadership_score * leadership_weight) + \
+                    (management_score * management_weight) + \
+                    (skill_match_score * skills_weight) + \
+                    (role_score * role_weight)
+    return round(overall_match, 2)
+
 def process_resume(resume, job_desc, progress_callback):
     resume_text = extract_text_from_file(resume.name)
     
@@ -127,23 +182,20 @@ def process_resume(resume, job_desc, progress_callback):
             "Candidate Name": "N/A",
             "Email": "N/A",
             "Contact": "N/A",
-            "Overall Match Percentage": "0%",
+            "Overall Match Percentage": 0.0,
             "Gemini Analysis": "Failed to extract text from resume."
         }
     
     try:
         gemini_analysis = analyze_with_gemini(resume_text, job_desc)
+        leadership_years, management_years, skills = extract_management_details(gemini_analysis)
+        role_keywords = gemini_analysis.lower()
+        overall_match = calculate_advanced_match(leadership_years, management_years, skills, required_skills, role_keywords)
         name, email, contact = extract_candidate_details(gemini_analysis)
-        
-        # Extract the match percentage directly from Gemini response
-        match_percentage_pattern = r"Overall Match Percentage\s*[:\-]?\s*(\d+)%"
-        match_percentage_match = re.search(match_percentage_pattern, gemini_analysis)
-        match_percentage = match_percentage_match.group(1) if match_percentage_match else "0"
-        
     except Exception as e:
         gemini_analysis = f"Gemini analysis failed: {str(e)}"
         name, email, contact = "N/A", "N/A", "N/A"
-        match_percentage = "0"
+        overall_match = 0.0
 
     progress_callback(1)  # Update progress for this resume
 
@@ -152,7 +204,7 @@ def process_resume(resume, job_desc, progress_callback):
         "Candidate Name": name,
         "Email": email,
         "Contact": contact,
-        "Overall Match Percentage": f"{match_percentage}%",
+        "Overall Match Percentage": f"{overall_match}%",
         "Gemini Analysis": gemini_analysis
     }
 
@@ -174,6 +226,9 @@ def analyze_resumes(resumes, job_desc):
     resume_count_message = f"{len(resumes)} resume(s) uploaded."
     return pd.DataFrame(results), resume_count_message
 
+def download_results(results):
+    return results.to_csv(index=False)
+
 # Gradio Interface with Submit Button and Progress Bar
 iface = gr.Interface(
     fn=analyze_resumes,
@@ -181,21 +236,12 @@ iface = gr.Interface(
         gr.File(label="Upload Resumes (PDF, DOCX, TXT)", file_count="multiple"),
         gr.Textbox(label="Job Description", lines=5)
     ],
-    outputs=[
-        gr.Dataframe(headers=["Resume", "Candidate Name", "Email", "Contact", "Overall Match Percentage", "Gemini Analysis"]),
-        gr.Textbox(label="Status Message")
-    ],
-    live=True,
-    title="Resume Analyzer with Leadership and Management Focus",
-    description="Upload resumes and a job description to calculate match percentages based on leadership, management, and skills.",
-    allow_flagging="never",
-    theme="default"
+    outputs=[gr.Dataframe(), gr.Textbox()],
+    live=False,  # Disable auto-running during input
+    allow_flagging="never"
 )
 
-# Add download option for the DataFrame
-def download_results(results_df):
-    return results_df.to_csv(index=False)
-
+# Add the file download option to the interface
 iface.add_component(gr.File(label="Download Results", file_output=download_results, visible=True))
 
-iface.launch(debug=True)
+iface.launch()