finance_agent / app.py
TusharLNT1
Initial commit
c30d6e8
import gradio as gr
import groq
import os
import json
from typing import Dict, List
import pandas as pd
from datetime import datetime
from dotenv import load_dotenv
# Load environment variables
load_dotenv()
class FinanceAIAgent:
def __init__(self, api_key: str):
self.client = groq.Client(api_key=api_key)
self.model = "llama-3.3-70b-versatile"
self.conversation_history = []
def generate_response(self, prompt: str, context: str = "") -> str:
# Combine context and prompt
full_prompt = f"{context}\n\nUser: {prompt}\nAssistant:"
try:
chat_completion = self.client.chat.completions.create(
model=self.model,
messages=[{"role": "user", "content": full_prompt}],
temperature=0.7,
max_tokens=1000
)
return chat_completion.choices[0].message.content
except Exception as e:
return f"Error generating response: {str(e)}"
def analyze_portfolio(self, portfolio_data: str) -> str:
prompt = f"""Analyze the following investment portfolio and provide insights:
{portfolio_data}
Include:
1. Risk assessment
2. Diversification analysis
3. Recommendations for rebalancing
4. Potential areas of concern"""
return self.generate_response(prompt)
def financial_planning(self, income: float, expenses: List[Dict], goals: List[str]) -> str:
prompt = f"""Create a financial plan based on:
Income: ${income}
Monthly Expenses: {json.dumps(expenses, indent=2)}
Financial Goals: {json.dumps(goals, indent=2)}
Provide:
1. Budget breakdown
2. Savings recommendations
3. Investment strategies
4. Timeline for achieving goals"""
return self.generate_response(prompt)
def market_analysis(self, ticker: str, timeframe: str) -> str:
prompt = f"""Provide a detailed market analysis for {ticker} over {timeframe} timeframe.
Include:
1. Technical analysis perspectives
2. Fundamental factors
3. Market sentiment
4. Risk factors
5. Potential catalysts"""
return self.generate_response(prompt)
def create_finance_ai_interface():
agent = FinanceAIAgent(api_key=os.getenv("GROQ_API_KEY"))
with gr.Blocks(title="Finance AI Assistant") as interface:
gr.Markdown("# Finance AI Assistant")
with gr.Tab("Portfolio Analysis"):
portfolio_input = gr.Textbox(
label="Enter portfolio details (ticker symbols and allocations)",
placeholder="AAPL: 25%, MSFT: 25%, GOOGL: 25%, AMZN: 25%"
)
portfolio_button = gr.Button("Analyze Portfolio")
portfolio_output = gr.Textbox(label="Analysis Results")
portfolio_button.click(
fn=agent.analyze_portfolio,
inputs=[portfolio_input],
outputs=portfolio_output
)
with gr.Tab("Financial Planning"):
with gr.Row():
income_input = gr.Number(label="Monthly Income ($)")
with gr.Row():
expenses_input = gr.Dataframe(
headers=["Category", "Amount"],
datatype=["str", "number"],
label="Monthly Expenses"
)
goals_input = gr.Textbox(
label="Financial Goals (one per line)",
placeholder="1. Save for retirement\n2. Buy a house\n3. Start a business"
)
planning_button = gr.Button("Generate Financial Plan")
planning_output = gr.Textbox(label="Financial Plan")
def process_financial_plan(income, expenses_df, goals):
expenses = expenses_df.to_dict('records')
goals_list = [g.strip() for g in goals.split('\n') if g.strip()]
return agent.financial_planning(income, expenses, goals_list)
planning_button.click(
fn=process_financial_plan,
inputs=[income_input, expenses_input, goals_input],
outputs=planning_output
)
with gr.Tab("Market Analysis"):
with gr.Row():
ticker_input = gr.Textbox(label="Stock Ticker")
timeframe_input = gr.Dropdown(
choices=["1 day", "1 week", "1 month", "3 months", "1 year"],
label="Timeframe"
)
market_button = gr.Button("Analyze Market")
market_output = gr.Textbox(label="Market Analysis")
market_button.click(
fn=agent.market_analysis,
inputs=[ticker_input, timeframe_input],
outputs=market_output
)
with gr.Tab("AI Chat"):
chatbot = gr.Chatbot()
msg = gr.Textbox(label="Ask anything about finance")
clear = gr.Button("Clear")
def respond(message, history):
history.append((message, agent.generate_response(message)))
return "", history
msg.submit(respond, [msg, chatbot], [msg, chatbot])
clear.click(lambda: None, None, chatbot, queue=False)
return interface
# Launch the interface
if __name__ == "__main__":
interface = create_finance_ai_interface()
interface.launch()
# import gradio as gr
# import groq
# import pandas as pd
# from datetime import datetime
# import plotly.express as px
# import json
# import os
# from typing import List, Dict
# from dotenv import load_dotenv
# # Load environment variables
# load_dotenv()
# # Initialize Groq client
# client = groq.Groq(api_key=os.environ["GROQ_API_KEY"])
# class FinanceAgent:
# def __init__(self):
# self.transactions = []
# self.budgets = {}
# self.goals = []
# def get_ai_advice(self, query: str) -> str:
# """Get financial advice from LLaMA model via Groq"""
# chat_completion = client.chat.completions.create(
# messages=[{
# "role": "system",
# "content": "You are a financial advisor. Provide clear, actionable advice."
# }, {
# "role": "user",
# "content": query
# }],
# model="llama-3.3-70b-versatile",
# temperature=0.7,
# )
# return chat_completion.choices[0].message.content
# def add_transaction(self, amount: float, category: str, description: str) -> Dict:
# """Add a new transaction"""
# transaction = {
# "date": datetime.now().strftime("%Y-%m-%d"),
# "amount": amount,
# "category": category,
# "description": description
# }
# self.transactions.append(transaction)
# return {"status": "success", "message": "Transaction added successfully"}
# def set_budget(self, category: str, amount: float) -> Dict:
# """Set a budget for a category"""
# self.budgets[category] = amount
# return {"status": "success", "message": f"Budget set for {category}"}
# def get_spending_analysis(self) -> Dict:
# """Analyze spending patterns"""
# df = pd.DataFrame(self.transactions)
# if df.empty:
# return {"status": "error", "message": "No transactions found"}
# spending_by_category = df.groupby('category')['amount'].sum().to_dict()
# return {
# "status": "success",
# "spending": spending_by_category,
# "total": sum(spending_by_category.values())
# }
# def create_interface():
# agent = FinanceAgent()
# with gr.Blocks(title="Personal Finance Assistant") as interface:
# gr.Markdown("# Personal Finance Assistant")
# with gr.Tab("Transactions"):
# with gr.Row():
# amount_input = gr.Number(label="Amount")
# category_input = gr.Dropdown(
# choices=["Groceries", "Utilities", "Entertainment", "Transportation", "Other"],
# label="Category"
# )
# description_input = gr.Textbox(label="Description")
# add_btn = gr.Button("Add Transaction")
# transaction_output = gr.JSON(label="Result")
# add_btn.click(
# fn=agent.add_transaction,
# inputs=[amount_input, category_input, description_input],
# outputs=transaction_output
# )
# with gr.Tab("Budgeting"):
# with gr.Row():
# budget_category = gr.Dropdown(
# choices=["Groceries", "Utilities", "Entertainment", "Transportation", "Other"],
# label="Category"
# )
# budget_amount = gr.Number(label="Budget Amount")
# set_budget_btn = gr.Button("Set Budget")
# budget_output = gr.JSON(label="Result")
# set_budget_btn.click(
# fn=agent.set_budget,
# inputs=[budget_category, budget_amount],
# outputs=budget_output
# )
# with gr.Tab("Analysis"):
# analyze_btn = gr.Button("Analyze Spending")
# spending_output = gr.JSON(label="Spending Analysis")
# analyze_btn.click(
# fn=agent.get_spending_analysis,
# outputs=spending_output
# )
# with gr.Tab("AI Advisor"):
# query_input = gr.Textbox(label="Ask for financial advice")
# advice_btn = gr.Button("Get Advice")
# advice_output = gr.Textbox(label="AI Advice")
# advice_btn.click(
# fn=agent.get_ai_advice,
# inputs=query_input,
# outputs=advice_output
# )
# return interface
# if __name__ == "__main__":
# interface = create_interface()
# interface.launch()