Welcome to AI-Trader Discussions!

[TianyuFan0504]

👋 Welcome!

We’re using Discussions as a place to connect with other members of our community. We hope that you:

• Ask questions you’re wondering about.
• Share ideas.
• Engage with other community members.
• Welcome others and are open-minded. Remember that this is a community we
  build together 💪.

To get started, comment below with an introduction of yourself and tell us about what you do with this community.

[Rgroenewold21]

Question: Are you considering the cost of trade when making a trade (in or out trade)?

[milliyang]

Question: How to ensure the a LLM model hav't learn a market value (look ahead bias) when you backtest ?

[doanbactam]

Not sure if you include transaction costs, API Key AI

[abacusus]

QuestIon: which type of trade is taken ?

[waveydev]

does it use fractional shares ?

[yifeng0223-gif]

可以设计一个docker容器安装的版本么？方便用nas或是其他软路由跑

[ABizoman]

It says in the readme that ALPHA advantage is used for market research, but I can only find Jina being used in the code.
Unless I'm missing something?

[bloomshao]

In MCP search ，"search": {"script": os.path.join(mcp_server_dir, "tool_alphavantage_news.py"), "name": "Search", "port": self.ports["search"]},

[ashuyagami-sketch]

if we can use a logics like a neural network which has wieghts based on chart or graph of past of the companies then it would transfer to a weight where it would check the price of trade then our money which we wil use for trade then we can make a good algorithm which would work perfectly seeing the circumstances but there will be bias which i can't make a scinario so please help me with it and i am working on intraday trading (dijkstra algorithm like thing finding the less weightd by adding the graph and price of trade)

[abacusus]

using a weight based logic would probably not work in this field becuase of the volatile market specially not in intraday , i would suggest using candle patterns logics which are used in trades like intraday.

[ashuyagami-sketch]

thanks for ur suggestion i am newbie in these work so it will take some time

[Cesar1019FV]

Are there plans to extend the system's operation to minerals (e.g., gold, silver)?

These assets typically exhibit upward trends, and determining the extent to which AI can enhance the trading margins would be valuable. If this is not currently in your roadmap, I would be interested in collaborating on the implementation of an AI trading system for minerals.

[ashuyagami-sketch]

lets try to get on work btw i am an 12 th grader in india so theres maximum chance i would be inactive for long time in mid periods but i will try to study it properly and analyze how it would work in these cases and how favourable it gets with time

[akshatrathee]

Can we work on work on a fork for India?

I would need help developing it but can potentially use this repo along with https://zerodha.com/z-connect/featured/connect-your-zerodha-account-to-ai-assistants-with-kite-mcp

[ashuyagami-sketch]

Can we briefly talk about this as I am new to it

[presser79]

Hi team,

I'm running a trading bot that uses the Market Intelligence API endpoint:
GET /api/market-intel/stocks/{symbol}/latest

The current_price field is returning prices that are 1-2 days old instead of current market prices.

Example tested on April 17, 2026:

• HD: API returns $338.91 (as_of: April 15) vs actual $350.34 → 3.26% off
• BA: API returns $218.88 (as_of: April 16) vs actual $226.76 → 3.48% off
• CAT: API returns $772.66 (as_of: April 16) vs actual $799.66 → 3.38% off
• AAPL: API returns $263.40 (as_of: April 16) vs actual $271.19 → 2.87% off

This causes two problems:

1. Bots record incorrect entry prices when trading
2. Followers cannot replicate signals at the displayed price

Could the API provide same-day prices, or add a clear price_as_of field so bots can detect stale data?

Related: #185

Thanks!