Create app.py

app.py

@@ -0,0 +1,42 @@
+from flask import Flask, request, jsonify
+from flask_cors import CORS
+import joblib
+import pandas as pd
+import os
+
+app = Flask(__name__)
+CORS(app)
+
+# In Spaces, files are in the same directory
+model = joblib.load('alumni_match_model.joblib')
+model_columns = joblib.load('model_feature_columns.joblib')
+
+@app.route('/', methods=['POST'])
+def handler():
+    incoming_data = request.get_json()
+    df = pd.DataFrame([incoming_data])
+
+    def count_common_skills(row):
+        viewer_skills = set(str(row.get('viewer_skills', '')).lower().split('|'))
+        target_skills = set(str(row.get('target_skills', '')).lower().split('|'))
+        return len(viewer_skills.intersection(target_skills))
+
+    df['common_skills_count'] = df.apply(count_common_skills, axis=1)
+    df['branch_match'] = (df['viewer_branch'].str.lower() == df['target_branch'].str.lower()).astype(int)
+
+    for col in model_columns:
+        if col.startswith('company_'):
+            df[col] = 0
+
+    company_name = incoming_data.get('target_company', '')
+    if company_name:
+        company_col_name = f"company_{company_name}"
+        if company_col_name in df.columns:
+            df[company_col_name] = 1
+
+    final_df = df[model_columns]
+    prediction_proba = model.predict_proba(final_df)
+    match_probability = prediction_proba[0][1]
+    final_score = round(match_probability * 10)
+
+    return jsonify({'score': final_score})