I built an AI that trades oil. It costs $0 to run. No subscriptions. No cloud bills. No excuses.
Most "AI trading bots" are a black box wrapped in a subscription fee. I got tired of that, so I built the opposite.
OIL-NEWS-SYSTEM is a complete, end-to-end autonomous pipeline that monitors real-time oil news, runs sentiment analysis against a hard-coded vocabulary of geopolitical and supply-chain signals, and executes trades on Binance Futures—without a human touching it.
Here is the architecture, stripped of the marketing:
🧠Sentiment Engine (sentiment.py): Not a generic LLM wrapper. A purpose-built vocabulary of bullish and bearish phrases scraped from months of real headlines—"production cut," "pipeline attack," "strait closed," "payrolls beat," "inventories draw." It classifies every tweet as BULL, BEAR, or NEUTRAL before a single dollar moves.
📡 Twitter Monitor (twitter_monitor.py): Polls a target account every 20 seconds. It primes on the first poll—records the baseline tweet ID and does not trade on stale news. It only fires when a genuinely new tweet appears.
⚡ Execution Layer (main.py): Connects to Binance Futures (BZUSDT), manages leverage, position sizing, stop-loss, and take-profit. It handles rate limits, IP bans, and reconnection logic. It runs as a Docker container on a Python 3.11 runtime.
The cost to run it: $0. No paid APIs in the critical path. No managed cloud services. A single docker run and it's live.
I also have an AI-generated video walking through the system—seeing it in motion is different from reading the code.
This is not a backtest. This is not a paper-trading demo. This is a working system built from scratch by someone who ships.
If you're building in the AI trading space and need an execution layer that actually works—or if you want to see what a real pipeline looks like before you hire someone to build yours—the code is public.
đź”— GitHub: https://github.com/S-7-D-2023/OIL-NEWS-SYSTEM
Ask me anything about the architecture. I'll answer with specifics, not buzzwords.
#AITrading #Python #Automation #FinTech #BuildInPublic #QuantDev #AI #oil
