Update app.py

app.py

@@ -1,14 +1,97 @@
+import os
 import gradio as gr
+import faiss
+import torch
+from sentence_transformers import SentenceTransformer
+from transformers import pipeline
 
-def greet(name):
-    return f"Hello {name}, your AI Twin is under construction!"
+# -----------------------------
+# Load Documents (Real + Synthetic)
+# -----------------------------
 
-iface = gr.Interface(
-    fn=greet,
+def load_documents():
+    docs = []
+    
+    # Load real KB
+    for file in os.listdir("knowledge_base"):
+        with open(f"knowledge_base/{file}", "r", encoding="utf-8") as f:
+            docs.append(f.read())
+    
+    # Load synthetic data
+    for file in os.listdir("synthetic_data"):
+        with open(f"synthetic_data/{file}", "r", encoding="utf-8") as f:
+            docs.append(f.read())
+    
+    return docs
+
+documents = load_documents()
+
+# -----------------------------
+# Create Embeddings
+# -----------------------------
+
+embed_model = SentenceTransformer("all-MiniLM-L6-v2")
+embeddings = embed_model.encode(documents)
+
+dimension = embeddings.shape[1]
+index = faiss.IndexFlatL2(dimension)
+index.add(embeddings)
+
+# -----------------------------
+# Load Generation Model (HF)
+# -----------------------------
+
+generator = pipeline(
+    "text-generation",
+    model="distilgpt2",
+    device=0 if torch.cuda.is_available() else -1
+)
+
+# -----------------------------
+# Retrieval Function
+# -----------------------------
+
+def retrieve(query, k=3):
+    query_embedding = embed_model.encode([query])
+    distances, indices = index.search(query_embedding, k)
+    return [documents[i] for i in indices[0]]
+
+# -----------------------------
+# RAG Function
+# -----------------------------
+
+def generate_answer(query):
+    context = retrieve(query)
+    
+    prompt = f"""
+    You are Hari's AI Twin.
+    Answer ONLY using the context below.
+    If answer is not found, say: Information not found in profile.
+
+    Context:
+    {context}
+
+    Question: {query}
+
+    Answer:
+    """
+
+    output = generator(prompt, max_length=400, num_return_sequences=1)
+    return output[0]["generated_text"].split("Answer:")[-1].strip()
+
+# -----------------------------
+# Gradio UI
+# -----------------------------
+
+def chatbot(query):
+    return generate_answer(query)
+
+interface = gr.Interface(
+    fn=chatbot,
     inputs="text",
     outputs="text",
     title="Hari's AI Twin",
-    description="Ask me anything about my profile!"
+    description="Ask me anything about my professional journey."
 )
 
-iface.launch()
+interface.launch()