diff --git a/Summit.md b/Summit.md
index a4e16a0a..d8edeff4 100644
--- a/Summit.md
+++ b/Summit.md
@@ -13,6 +13,12 @@ Then you will receive 2 invitation to join the :
 - Github MLOpsPython team.
 - AzureML CatsDogsOthers project : https://ml.azure.com/home (you should see project name: Cats-Dogs)
 
+## 0. Prerequisite
+
+- Pycharm (https://www.jetbrains.com/pycharm/)
+- Download and install python 3.11.x on your laptop. 
+https://www.python.org/downloads/
+
 ## 1. Introduction
 
 [README.md](README.md)
diff --git a/packages/inference/src/mlopspython_inference/inference_pillow.py b/packages/inference/src/mlopspython_inference/inference_pillow.py
index 63e5a9d3..6dfa4c9d 100644
--- a/packages/inference/src/mlopspython_inference/inference_pillow.py
+++ b/packages/inference/src/mlopspython_inference/inference_pillow.py
@@ -1,4 +1,6 @@
+from dataclasses import dataclass
 from io import BytesIO
+from typing import List
 
 import numpy as np
 from tensorflow.keras.preprocessing.image import load_img, img_to_array
@@ -22,16 +24,20 @@ def load_image(filename: str|BytesIO):
 
 BASE_PATH = Path(__file__).resolve().parent
 
+@dataclass
+class InferenceOutput:
+    prediction: str
+    values: List[float]
 
 class Inference:
     def __init__(self, logging, model_path: str):
         self.logger = logging.getLogger(__name__)
         self.model = load_model(model_path)
 
-    def execute(self, filepath:str|BytesIO):
+    def execute(self, filepath:str|BytesIO) -> InferenceOutput:
         img = load_image(filepath)
         result = self.model.predict(img)
         values = [float(result[0][0]), float(result[0][1]), float(result[0][2])]
         switcher = ['Cat', 'Dog', 'Other']
         prediction = np.argmax(result[0])
-        return {"prediction": switcher[prediction], "values": values}
+        return InferenceOutput(switcher[prediction], values)
diff --git a/packages/inference/src/tests/inference_tests.py b/packages/inference/src/tests/inference_tests.py
index 96d3122a..fb013f41 100644
--- a/packages/inference/src/tests/inference_tests.py
+++ b/packages/inference/src/tests/inference_tests.py
@@ -11,12 +11,14 @@
 class TestInference(unittest.TestCase):
 
     def test_inference(self):
+        # Arrange
         inference = Inference(logging, str(input_directory / "model" / "final_model.h5"))
+        # Act
         inference_result = inference.execute(str(input_directory / "images" / "cat.png"))
-
+        # Assert
         expected_result = {'prediction': 'Cat', 'values': [1.0, 2.370240289845506e-30, 0.0]}
-        self.assertEqual(inference_result['prediction'], expected_result['prediction'])
-        self.assertEqual(len(inference_result['values']), len(expected_result['values']))
+        self.assertEqual(inference_result.prediction, expected_result['prediction'])
+        self.assertEqual(len(inference_result.values), len(expected_result['values']))
 
 
 if __name__ == "__main__":
diff --git a/production/api/core/model/inference_pillow.py b/production/api/core/model/inference_pillow.py
index b4f9c4db..9be2c6b9 100644
--- a/production/api/core/model/inference_pillow.py
+++ b/production/api/core/model/inference_pillow.py
@@ -1,3 +1,4 @@
+import dataclasses
 import io
 from pathlib import Path
 from mlopspython_inference.inference_pillow import Inference as InferencePillow
@@ -11,4 +12,5 @@ def __init__(self, logging, app_settings):
         self.inference = InferencePillow(logging, str(model_path))
 
     def execute(self, file, filename, settings=None):
-        return self.inference.execute(io.BytesIO(file.read()))
+        prediction = self.inference.execute(io.BytesIO(file.read()))
+        return dataclasses.asdict(prediction)
diff --git a/step_0_setup.md b/step_0_setup.md
index b3df03c2..16c1a015 100644
--- a/step_0_setup.md
+++ b/step_0_setup.md
@@ -1,6 +1,6 @@
 # Setup
 
-## 1. Download ans install python 3.11.x on your laptop. 
+## 1. Download and install python 3.11.x on your laptop. 
 https://www.python.org/downloads/
 
 It is a good practice to always use the lastest version with supports: 
diff --git a/train/evaluate/evaluate.py b/train/evaluate/evaluate.py
index 57b253d6..6e5b7ef6 100644
--- a/train/evaluate/evaluate.py
+++ b/train/evaluate/evaluate.py
@@ -21,7 +21,7 @@ def evaluate(logging, input_model_directory: Path, input_images_directory: Path,
             continue
         model_result = model.execute(str(path))
 
-        prediction = model_result["prediction"]
+        prediction = model_result.prediction
         prediction_truth = path.parent.name.lower().replace("s", "")
         status = prediction_truth == prediction.lower()
         statistics["ok" if status else "ko"] += 1
@@ -29,7 +29,7 @@ def evaluate(logging, input_model_directory: Path, input_images_directory: Path,
                   "ok": status,
                   "prediction": prediction,
                   "prediction_truth": prediction_truth,
-                  "values": model_result["values"]}
+                  "values": model_result.values}
         results.append(result)
     statistics["total"] = statistics["ok"] + statistics["ko"]