Upload 2 files

app.py

@@ -0,0 +1,105 @@
+import streamlit as st
+from phi.agent import Agent
+from phi.model.google import Gemini
+from phi.tools.duckduckgo import DuckDuckGo
+from google.generativeai import upload_file, get_file
+import google.generativeai as genai
+import time
+from pathlib import Path
+import tempfile
+from dotenv import load_dotenv
+load_dotenv()
+import os
+
+API_KEY = os.getenv("GOOGLE_API_KEY")
+if API_KEY:
+    genai.configure(api_key = API_KEY)
+
+# Page Configuration
+st.set_page_config(
+    page_title="Multimodal AI Agent - Video Summarizer",
+    page_icon="(1f4f9_videocamera) ",
+    layout="wide"
+)
+
+st.title("Phidata Multimodal AI Summarizer Agent (1f4f9_videocamera) ")
+st.header("Powered by Gemini 2.0 Flash Exp")
+
+@st.cache_resource
+def initialize_agent():
+    return Agent(
+        name = "Video AI Summarizer",
+        model = Gemini(id="gemini-2.0-exp"),
+        tools = [DuckDuckGo()],
+        markdown = True
+    )
+
+# Initialize Agent 
+multimodal_agent = initialize_agent()
+
+# File uploader
+video_file = st.file_uploader(
+    "Upload a video file", type=['mp4', 'mov', 'avi'], help="Upload a video for AI Analysis"
+)
+if video_file:
+    with tempfile.NamedTemporaryFile(delete=False, suffix='.mp4') as temp_video:
+        temp_video.write(video_file.read())
+        video_path = temp_video.name
+
+    st.video(video_path, format="video/mp4", start_time=0)
+
+    user_query = st.text_area(
+        "What insights are you seeking from the video?",
+        placeholder="Ask anything about the video content. The AI Agent will analyze and gather additional information",
+        help="Provide specific question or insights you want from the video."
+    )
+
+    if st.button("Analyze video", key="analyze_video_button"):
+        if not user_query:
+            st.warning("Please enter a question or insight to analyze the video.")
+        else:
+            try:
+                with st.spinner("Processing video and gathering insights..."):
+                    #Upload and process video file
+                    processed_video = upload_file(video_path)
+                    while processed_video.state.name == "PROCESSING":
+                        time.sleep(1)
+                        processed_video = get_file(processed_video.name)
+
+                    #prompt generation for analysis
+                    analyses_prompt = (f"""
+                    Analyze the uploaded video for content and context.
+                    Respond the following query using the video insights and supplimentary web research
+                    {user_query}
+
+                    Provide a detailed, user-friendly and actionable response.
+                        """
+                    )
+
+                    #AI agent processing 
+                    response = multimodal_agent.run(analyses_prompt, video = [processed_video])
+
+                #Display the result 
+                st.subheader("Analysis result")
+                st.markdown(response.content)
+            
+            except Exception as error:
+                st.error(f"An error occured during analysis: {error}")
+            finally:
+                # Clean temporary file
+                Path(video_path).unlink(missing_ok=True)
+else:
+    st.info("Upload a video file to begin analysis")
+
+# Customize text are heihght
+
+st.markdown(
+    """
+    <style>
+    .stTextArea textarea {
+        height: 100px;
+    }    
+    </style>
+    """,
+    unsafe_allow_html=True
+)

requirements.txt

@@ -0,0 +1,5 @@
+streamlit
+phidata
+google-generativeai
+openai
+duckduckgo-search