Divine Signal
Crypto trading signals from Telegram channels, read automatically, turned into structured data, and followed until they hit a target or a stop loss.
Visit the site (opens in a new tab)- For
- Following crypto trading signals
- What it is
- A web app that reads Telegram messages
- Our part
- All of it; this is our own product
- Built with
- Python, FastAPI, PostgreSQL, React
Open the full-size image (opens in a new tab)The need
Telegram channels post trading signals as free-form messages, each in its own format. They're hard to compare, let alone to check which ones actually hit their targets.
Divine Signal reads those messages, turns them into tidy data, and follows the market price for each signal until there's an outcome.
What we built
A page for every signal
Each signal has its own page: a price chart, the entry, target, and stop-loss levels, and its history.
- Price chart with the entry, targets (TP), and stop loss (SL).
- Targets that were hit are recorded with the time.
- The signal's history, from creation to the last target hit.
Outcome statistics
Outcomes can be compared by channel, by coin, or by direction. Signals that never filled or expired count as neutral, not as losses.
Open the full-size image (opens in a new tab)- Compare by channel, by coin, or by direction.
- A summary of every signal that has finished.
- The share of targets hit, stops hit, and neutral signals in each group.
Also included
- Admin review queue
- Source channel management
- User management
- Activity history
From message to outcome
Message arrives
From a Telegram channel being followed.
Read and checked
Turned into data with a strict format, then checked for anything odd.
Reviewed if odd
Doubtful signals go to an admin queue and stay hidden for now.
Shown in the feed
It appears in the browser right away, without reloading the page.
Followed to the end
Checked against minute-by-minute prices until it hits a target or the stop loss.
Details we took care of
- Doubt first, show later
- Messages that don't match the format or look odd aren't published straight away. An admin reviews them first.
- Outcomes from minute prices
- Targets and stop losses are checked against each minute's high and low. If both are hit in the same minute, the signal counts as stopped out.
- Never recorded twice
- The same message is never stored twice, and a signal reposted within 10 minutes is recognized as the same signal.
- Live even when the connection drops
- The feed reconnects on its own and catches up on missed updates, without showing a signal twice.
The same approach works for
- Orders from chats turned into a tidy list
- Field reports from WhatsApp or Telegram
- Dashboards that update themselves
- Price or stock monitoring
Notes
Divine Signal is our own product. Source channel names are blurred.
Percentages and statistics in the app are records of signals, not verified trading returns, and not investment advice.
More projects
- WiFi Billing
- Monthly billing for neighborhood internet providers. Invoices go out on their own, customers get reminders on WhatsApp, and cashiers check payment receipts in one place.
- Affluena
- Personal finance for individuals and couples, on web and mobile. Wallets can be shared, so two people work from the same numbers.
- SIPPKBPD
- Electronic ID card recording requests for residents with disabilities. Each request moves up from the neighborhood to the district, and every step is recorded.
- CLC Readykit
- A web app for a Mandarin school for children: from parents' inquiries and lead follow-ups to classes, invoices, and payments.
Need something like this?
Tell us what you need over WhatsApp. The first conversation is free, and if something simpler would do, we'll say so.