Streaming Analytics for the Digital Asset Risk Management System - Cloudwall Success Story

About the Client

Cloudwall (part of Talos now) is a fintech company that provides institutional-grade risk management for digital asset markets. Their platform helps clients monitor exposure across crypto spot and derivatives markets by analyzing price data, volatility, and counterparty risk in real time. Cloudwall’s competitive edge depends on processing vast amounts of market data quickly, accurately, and cost-effectively.

When the Data Streams Get too Fast to Handle

As trading volumes surged and Cloudwall expanded its coverage across new markets and tokens, its data pipeline struggled to keep pace. Processing over 100 million messages per hour from multiple market feeds introduced latency, inconsistency, and rising operational costs. Onboarding new tokens often meant manual reconfiguration and downtime. Cloudwall needed a real-time data architecture that could keep up with volatility, scale seamlessly, and support reliable pricing and risk models—without driving infrastructure overhead out of control.

A Streaming Analytics Platform Powered by Flink

Xebia’s team (formerly GetInData) worked closely with Cloudwall to build a robust, cloud-native streaming analytics system based on Apache Flink. The platform ingests, harmonizes, and enriches real-time market feeds from multiple providers. Fair price metrics are calculated dynamically, and Redis-backed caches eliminate redundant external API calls—drastically reducing cloud egress costs.

Apache Kafka serves as the backbone for scalable, high-throughput messaging. Flink's dynamic configuration allows Cloudwall to onboard new tokens or adjust logic without system downtime. The architecture is containerized and deployed in Kubernetes, ensuring security, observability, and fast iteration cycles. The solution gives Cloudwall a unified foundation to drive fast, automated decision-making in volatile markets.

The Results

Cloudwall now processes over 100 million messages per hour with millisecond-level latency, enabling real-time risk monitoring, accurate pricing, and timely alerts. Operational costs have been reduced thanks to Redis caching and optimized infrastructure. New tokens can be added on the fly—without downtime or engineering bottlenecks—allowing Cloudwall to expand rapidly and adapt to market shifts. With this platform in place, Cloudwall delivers more accurate insights, faster, and with greater confidence to institutional clients navigating digital asset risk.