ES Futures Alert Bot
AI extracts trading levels from tweets. Telegram delivers them in under 60 seconds. Built start-to-finish in a single day.
Full system: scraping, AI extraction, Telegram delivery, price alerts
From tweet to Telegram notification
From Tweet to Trade
Trading tweets contain valuable levels, but they're buried in unstructured text. 'Supports at 35/20, targeting 70/90, below 10 look for 5980' means nothing to machines. Traders miss updates or misread shorthand. By the time you parse the tweet, the trade's already moved.
"The solution had to understand trading shorthand, resolve ambiguous prices from context, and deliver structured data fast enough to matter."
Trading levels extracted and delivered
ES futures market hours coverage (Sun-Fri)
A Telegram bot that scrapes trading tweets, uses AI to extract structured price levels (supports, targets, fail levels), and delivers formatted alerts to subscribers. When ES futures approach extracted levels, users get price alerts with visual charts showing the levels overlaid on price data.
AI Level Extraction
Custom-tuned prompts parse trading tweets into structured data. Handles shorthand like '35' meaning 6935 based on current ES price context. Extracts supports, targets, fail levels, downside targets, and reclaim levels.
- • Parses trading shorthand and abbreviations
- • Resolves ambiguous prices from market context
- • Structured JSON output for every tweet
Real-Time Telegram Alerts
New trading levels hit Telegram within seconds of the original tweet. Formatted messages show the full picture: bias, supports, targets, fail levels, and context. No more refreshing Twitter.
- • Sub-60-second tweet-to-alert delivery
- • Formatted messages with full level breakdown
- • Subscription tiers with multiple trader slots
Visual Price Charts
When ES approaches an extracted level, the bot sends a chart showing price action with all levels overlaid. See exactly where you are relative to supports and targets.
- • Auto-generated charts on price approach
- • All levels visualized on price data
- • Market hours awareness (pre-market, RTH, overnight)
Built in a Day
Full system built and deployed in under 24 hours
Tweet-to-Telegram delivery in under 60 seconds
What I Learned
AI prompt engineering is the real work. Getting consistent structured output from unstructured trading tweets took more iteration than the code itself.
Telegram Bots are underrated. The API is clean, webhooks work reliably, and users already have the app installed.
Speed matters in trading tools. Every minute of latency is a minute traders are behind. Optimize the pipeline, not just the code.
Need trading tools or Telegram bots?
Custom bots, alert systems, AI integrations. If it involves extracting signal from noise, I've probably built something similar.
Get in touch
Yixn.io
ES Futures Bot
Hester NG
SparkChambers
Kofferly
Anti-Baby-Shower
SipQuest
PintRush
Linkster.co
Golem Overlord
Splex.gg
ProPhy