Skip to main content
All case studies
Fintech

High-frequency trading dashboard with sub-50ms updates

Client: Meridian Capital

Challenge

Meridian's traders needed a real-time dashboard processing market data from 12 exchanges with sub-50ms latency. Their legacy React app caused frame drops and memory leaks during high-volatility periods.

Solution

We rebuilt the frontend with Next.js 15 App Router, WebSocket streams via PartyKit, and a custom canvas-based order book visualization. Backend migrated to Go with Redis Streams for event sourcing.

Architecture

Event-driven architecture with Redis Streams as the message bus. Next.js renders server components for static chrome, while client components handle WebSocket data streams. Go microservices process normalized market data and push deltas to edge workers.

Stack

Next.js 15TypeScriptGoRedisWebSocketsPartyKitTimescaleDB

Key Metrics

UI update latency
<50ms
Memory reduction
-78%
Frame rate
60fps
Uptime
99.99%

Outcomes

  • Eliminated frame drops during market volatility spikes
  • Reduced time-to-insight for traders by 40%
  • Cut infrastructure costs by 35% through edge caching
  • Achieved SOC 2 Type II compliance within 3 months