diff --git a/apps/profiles/serializers/institution_serializers.py b/apps/profiles/serializers/institution_serializers.py
index 318ccdc..ad44dc1 100644
--- a/apps/profiles/serializers/institution_serializers.py
+++ b/apps/profiles/serializers/institution_serializers.py
@@ -231,6 +231,68 @@ class InstitutionCreateUpdateSerializer(serializers.ModelSerializer):
             'is_featured',
         ]
 
+    def to_internal_value(self, data):
+        if hasattr(data, 'copy'):
+            data = data.copy()
+        else:
+            data = dict(data)
+
+        # Normalize institution type if passed in Persian or other aliases
+        inst_type = data.get('type')
+        if inst_type:
+            type_map = {
+                'مؤسسه فرهنگی و مذهبی': 'cultural_center',
+                'مرکز فرهنگی': 'cultural_center',
+                'مسجد': 'mosque',
+                'حسینیه': 'hussainiya',
+                'حوزه علمیه': 'institute',
+                'مرکز آموزشی': 'institute',
+                'مرکز پژوهشی': 'institute',
+                'مرکز خیریه': 'charity',
+                'خیریه': 'charity',
+                'کتابخانه': 'library',
+                'سایر': 'other',
+            }
+            if inst_type in type_map:
+                data['type'] = type_map[inst_type]
+
+        # Normalize established / established_year
+        est = data.get('established') if data.get('established') is not None else data.get('established_year')
+        if est is not None and not data.get('established_year'):
+            try:
+                persian_digits = '۰۱۲۳۴۵۶۷۸۹'
+                arabic_digits = '٠١٢٣٤٥٦٧٨٩'
+                est_str = str(est).strip()
+                for i, d in enumerate(persian_digits):
+                    est_str = est_str.replace(d, str(i))
+                for i, d in enumerate(arabic_digits):
+                    est_str = est_str.replace(d, str(i))
+                if est_str:
+                    data['established_year'] = int(est_str)
+            except (ValueError, TypeError):
+                data['established_year'] = None
+
+        # Normalize website: prepend https:// if missing scheme
+        website = data.get('website')
+        if website and isinstance(website, str) and website.strip():
+            website = website.strip()
+            if not (website.startswith('http://') or website.startswith('https://')):
+                data['website'] = f'https://{website}'
+
+        # Normalize camelCase socialMedia / workingHours if provided
+        if 'socialMedia' in data and not data.get('social_media'):
+            data['social_media'] = data.get('socialMedia') or {}
+        if 'workingHours' in data and not data.get('working_hours'):
+            data['working_hours'] = data.get('workingHours') or {}
+
+        # If cover_image or avatar is not an actual uploaded file (e.g. string path or URL), ignore it so ImageField doesn't fail
+        for img_field in ['cover_image', 'avatar']:
+            val = data.get(img_field)
+            if val is not None and not hasattr(val, 'read'):
+                data.pop(img_field, None)
+
+        return super().to_internal_value(data)
+
     def create(self, validated_data):
         user = self.context.get('request').user if self.context.get('request') else None
         if user and user.is_authenticated: