Market Commentary Assistant – Telegram Bot

A Telegram-based Investment Coach & Market Commentary Assistant built with FastAPI and Groq (OpenAI-compatible).

The assistant helps users:

• Learn investing concepts (Beginner → Intermediate)
• Get neutral, news-style market commentary

🚫 Crucial constraint: The bot never provides buy/sell recommendations, stock tips, intraday calls, or personalized investment advice. It acts strictly as a coach/commentator, not an advisor.

---

✨ Key Features

1. Investment Coach Mode

• Explains concepts like ETFs, SIPs, risk, diversification
• Uses simple language and analogies
• Educational only, no product or stock recommendations

2. Market Commentary Mode

• Answers questions like:
  • “Why is the market down today?”
  • “Summarize today’s market news”
• Neutral, journalist-style summaries
• Focus on macro, rates, earnings, global cues

3. Safety & Compliance Guardrails

• Refuses:
  • Stock picks
  • Buy/sell advice
  • Intraday tips
  • Guaranteed returns
• Always redirects users to educational frameworks
• Clear disclaimer included in relevant responses

---

🧠 System Design Overview

User (Telegram)
      ↓
Telegram Webhook
      ↓
FastAPI (Render)
      ↓
Mode Router (Coach / Commentary / Guardrail)
      ↓
Groq LLM
      ↓
Safe, compliant response

---

🛠 Tech Stack

Backend

• Python
• FastAPI (API + service layer)

LLM

• Groq API
• Model: llama-3.3-70b-versatile
• Chosen for:
  • Low cost / free-tier friendliness
  • Fast inference
  • Open-source model support

Interface

• Telegram Bot (Webhook mode)

Data

• Mock market news data

Infrastructure (Free-tier compatible)

• Render

---

📁 Project Structure

project-root/
│
├── app/
│   ├── main.py               # FastAPI entry point
│   ├── telegram_webhook.py   # Telegram webhook handler
│   ├── llm.py                # Groq client
│   ├── router.py             # Intent / mode detection
│   ├── prompts.py            # System prompts
│   ├── news.py               # Mock market news
│   └── safety.py             # Output safety enforcement
│
├── requirements.txt
├── .env.example
└── README.md

---

⚙️ Local Setup Instructions

1. Prerequisites

• Python 3.11+
• A Groq API key
• A Telegram Bot Token (via @BotFather)

---

2. Clone the Repository

git clone  https://github.com/SAKIB0004/Market_Assistant_Bot.git
cd Market_Assistant_Bot

---

3. Create & Activate Virtual Environment

python -m venv .venv

# Windows
.venv\Scripts\activate

# macOS / Linux
source .venv/bin/activate

---

4. Install Dependencies

pip install -r requirements.txt

---

5. Configure Environment Variables

Create a .env file in the project root:

GROQ_API_KEY="groq_api_key"
GROQ_MODEL=llama-3.3-70b-versatile
TELEGRAM_BOT_TOKEN="telegram_bot_token"

---

6. Run Locally (Webhook-style API)

uvicorn app.main:app --reload

---

🚀 Deployment on Render (Telegram Webhook)

1. Create a Render Web Service

• Environment: Python
• Build Command:

  pip install -r requirements.txt

• Start Command:

  uvicorn app.main:app --host 0.0.0.0 --port $PORT

---

Render URL:

https://market-assistant-bot.onrender.com

Using the Telegram Bot

The bot is deployed and available on Telegram.

How to Find the Bot

1. Open Telegram
2. Search for the bot by username: Market_Commentary_Assistant_Bot

or

🔗 Live Telegram Bot: https://t.me/Market_Commentary_Assistant_Bot

Start the Bot

1. Open the chat
2. Click Start
3. Send a message (no commands required)

---

🧪 Example Test Prompts

Coach Mode

• Explain ETFs
• What is SIP?
• How does diversification reduce risk?

Commentary Mode

• Why is the market down today?
• Summarize today’s market news

Guardrail Test

• Which stock should I buy today?
• Give me intraday tips

(Expected: refusal + educational redirection)

---

🔒 Compliance & Safety Notes

• No stock recommendations
• No buy/sell language
• No personalized advice
• Explicit refusal + redirection logic

This ensures the assistant remains a coach/commentator, not an investment advisor.

⚠️ Disclaimer

This project is for educational purposes only and does not constitute financial or investment advice. Investing involves risk, and users should perform their own research or consult a qualified professional before making financial decisions.

---

🙌 Acknowledgements

Built as part of an Investment Coach + Market Commentary Assistant take-home assignment, with a focus on safe, compliant GenAI system design.