Case Study · Development
Multi-exchange crypto data aggregation and real-time analytics with Discord bot integration.
Crypto traders watching multiple exchanges simultaneously face a fragmented picture — each platform has its own interface, its own delays, its own gaps. CryptoWatchTools was built to collapse that sprawl into a single aggregated stream: normalized data from every connected exchange, surfaced where traders already live.
The brief was to design and build a robust data pipeline capable of ingesting order-book and price data from multiple exchanges simultaneously, normalizing it into a consistent schema, and delivering real-time alerts and analytics through a Discord bot — without a web dashboard to maintain.
How it's built
The architecture centres on a fanout-capable ingestion core — new exchange adapters slot in without touching analytics or delivery logic.
Persistent WebSocket connections to each exchange, with automatic reconnect, heartbeat management, and per-adapter message parsing.
Every exchange speaks a different dialect. The normalization layer translates each into a single canonical schema — price, volume, side, timestamp — so analytics are exchange-agnostic.
Configurable signal detection: volume spikes, cross-exchange spread divergence, momentum triggers, and percentage-move alerts — all streamed as they happen.
Per-user symbol watchlists persisted in PostgreSQL. Users subscribe to specific pairs and receive only the signals they care about — no noise.
All incoming ticks are written to timestamped PostgreSQL tables, enabling backtesting queries, trend analysis, and historical comparisons on demand.
No web dashboard to maintain or log into. Alerts and query responses land directly in Discord channels, where traders are already coordinating.
Discord Integration
The bot is the interface. No separate app, no separate login — traders interact with live market data through slash commands in their existing Discord server.
Query live price, spread, volume, and order-book depth for any supported pair — from any channel the bot is in.
Subscribe to price-move, volume-spike, or spread-divergence alerts per symbol. Alerts fire to a designated channel as they trigger.
Each user's watchlist is persistent — subscribe once and your pairs follow you across sessions without reconfiguring.
The mark
The mark reads the product at a glance. Three ascending chart bars — short, medium, tall — show data accumulating in real time. An orbit ellipse wraps around them, suggesting surveillance: the market under constant observation. The crescent nods to crypto's lunar vocabulary without being literal about it.
Real-time integrations, custom analytics, and delivery where your team already works.
