Skip to content

Suryadharmaa/ATLAS

Repository files navigation

ATLAS — AI Market Co-Pilot

ATLAS is a Telegram-based AI market analysis assistant for Forex, Crypto, Metals, and selected stocks. It is designed as a co-pilot, not a signal bot: the goal is to help users understand market context, think through risk, and make more disciplined decisions.

What ATLAS does

  • Accepts market questions from Telegram chat or commands
  • Detects the user’s intent and language (ID/EN)
  • Sends the request to Groq for fast AI reasoning
  • Returns a structured analysis with:
    • Bias
    • Reasoning
    • Alternative scenario
    • Risks
    • Data strength
  • Adds a disclaimer so the bot stays educational and risk-aware

Core product principles

  1. Context over prediction
    ATLAS explains what is happening, not “guaranteed” market direction.

  2. Risk first
    Every answer should remind the user that market decisions carry risk.

  3. Simple MVP
    The first version is intentionally small so it can be tested quickly.

  4. Low-cost stack
    Built to work with free or low-budget tooling.

Current stack

The prototype uses:

  • python-telegram-bot for Telegram bot handling
  • openai SDK for Groq-compatible chat completions
  • python-dotenv for environment variables

How the bot works

User message in Telegram
→ bot.py receives the message
→ session.py stores language / history / warning state
→ analyzer.py detects ticker and builds prompts
→ groq_client.py calls Groq API
→ formatter.py cleans and validates the result
→ Telegram sends the final answer back to the user

Project structure

atlas-bot/
├── bot.py           # Telegram handlers and app entry point
├── analyzer.py      # Intent parsing and analysis workflow
├── groq_client.py   # Groq API client
├── prompts.py       # System prompt, user prompt, language helpers
├── formatter.py     # Output cleanup and safety checks
├── session.py       # In-memory chat session state
├── config.py        # Environment variables and constants
├── requirements.txt # Python dependencies
└── README.md        # Project documentation

Commands

  • /start — reset session and show onboarding
  • /help — show bot usage
  • /disclaimer — show the full disclaimer
  • /analyze [ticker] — run a market analysis

Example

/analyze BTC
/analyze XAUUSD
/analyze BBCA

You can also send free text such as:

analisa XAUUSD hari ini
analyze BTC
what is the bias on ETH?

Environment variables

Create a .env file with:

TELEGRAM_BOT_TOKEN=your_telegram_bot_token
GROQ_API_KEY=your_groq_api_key

Install

pip install -r requirements.txt

Run locally

python bot.py

Telegram setup

  1. Open @BotFather in Telegram
  2. Create a new bot
  3. Copy the bot token into .env
  4. If testing in a group, consider disabling privacy mode in BotFather so the bot can see normal messages

Analysis style

ATLAS is designed to always return analysis in a structured format:

  • Bias
  • Reasoning
  • Alternative Scenario
  • Risks
  • Data Strength

The bot should avoid:

  • guaranteed profit language
  • buy/sell commands
  • overconfident claims
  • unstructured replies

Disclaimer

ATLAS is an educational market analysis tool. It does not provide financial advice, investment guarantees, or trading promises. All decisions remain the user’s responsibility.

Roadmap idea

  • Phase 1: Telegram bot + structured AI analysis
  • Phase 2: Realtime market data integration
  • Phase 3: Memory, richer analytics, and web dashboard

Brand visual

A concept banner for ATLAS has been generated for product presentation.


Built for fast iteration, low budget, and high clarity.

About

Atlas Financial Bot

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors

Languages