diff --git a/api/app.py b/api/app.py
deleted file mode 100644
index 77baf21..0000000
--- a/api/app.py
+++ /dev/null
@@ -1,39 +0,0 @@
-from fastapi import FastAPI
-import pandas as pd
-import numpy as np
-from sentence_transformers import SentenceTransformer
-from sklearn.metrics.pairwise import cosine_similarity
-
-app = FastAPI()
-
-# Load the dataset
-dataset_url = "https://raw.githubusercontent.com/nehaprabhavalkar/indian-food-101/main/indian_food.csv"
-df = pd.read_csv(dataset_url)
-
-# Initialize SentenceTransformer model
-bert_model = SentenceTransformer("bert-base-nli-mean-tokens")
-
-
-# Function to calculate embeddings for recipes
-def calculate_embeddings(data):
-    embeddings = bert_model.encode(data).tolist()
-    return embeddings
-
-
-# Function to recommend similar recipes
-def recommend_similar_recipes(recipe_name, embeddings):
-    recipe_index = df[df["name"] == recipe_name].index[0]
-    similarities = cosine_similarity([embeddings[recipe_index]], embeddings)[0]
-    similar_indices = np.argsort(similarities)[::-1][1:11]  # Exclude self
-    similar_recipes = df.iloc[similar_indices]["name"].tolist()
-    return similar_recipes
-
-
-# API endpoint to recommend similar recipes
-@app.get("/recommend/{recipe_name}")
-async def recommend_recipe(recipe_name: str):
-    # Calculate embeddings for recipe ingredients
-    embeddings = calculate_embeddings(df["ingredients"])
-    # Recommend similar recipes
-    recommended_recipes = recommend_similar_recipes(recipe_name, embeddings)
-    return {"recommended_recipes": recommended_recipes}
diff --git a/vercel.json b/vercel.json
index 2160152..53a61a6 100644
--- a/vercel.json
+++ b/vercel.json
@@ -1,10 +1,14 @@
 {
-    "version": 2,
-    "routes": [
-      {
-        "src": "/(.*)",
-        "dest": "/app.py"
-      }
-    ]
-  }
-  
\ No newline at end of file
+  "builds": [
+    {
+      "src": "app.py",
+      "use": "@vercel/python"
+    }
+  ],
+  "routes": [
+    {
+      "src": "/(.*)",
+      "dest": "app.py"
+    }
+  ]
+}