What tech stack does Kraken use?

Kraken's most important architectural decision is its multi-year migration from PHP to Rust. The Core Backend team began rewriting services in Rust in mid-2018 and has accumulated 500,000+ lines of Rust code across hundreds of microservices, surpassing the PHP codebase in size and putting the legacy PHP services into a managed wind-down. Rust services use Tokio (async runtime) and handle 150,000 requests per second per instance at sub-3ms p99.9 latency — a performance envelope that the PHP codebase could not achieve. Service communication is standardized on gRPC across Rust, Go, C++, and Python services, reflecting Kraken's organic growth through multiple engineering generations rather than a single-language-from-scratch greenfield build.

Infrastructure is a hybrid of AWS (for elastic workloads, streaming, and developer tooling) and on-premises co-located data centers for the most latency-sensitive trading paths. Kraken Pro achieved round-trip latencies below 2ms using VPC peering between AWS and co-located DCs. The company runs Kubernetes for container orchestration (after a multi-year migration from Nomad), Terraform for infrastructure-as-code, ArgoCD for GitOps continuous delivery, and Kafka powering an in-house Rust-based stream-processing framework for real-time market data and order events. For high-availability fault-tolerant systems, Kraken uses Aeron multicast — a high-performance messaging library common in latency-sensitive financial infrastructure. Cloudflare sits in front of all public-facing endpoints for DDoS protection and global CDN distribution; NGINX handles internal reverse proxying.

Job listings active through early 2026 confirm continued investment in Kubernetes expertise, Kafka or equivalent event-streaming experience, and Rust proficiency as baseline requirements for senior backend roles — signals that the core architectural choices are stable rather than in flux.

Kraken's user-facing web products are built with React and Next.js, with TypeScript as the primary frontend language. The Kraken engineering blog (engineering.kraken.tech) appears to run a Django backend for internal tooling and supporting services — consistent with Django's use at several fintech companies for admin surfaces and internal dashboards. Bootstrap appears in older frontend surfaces and is likely being phased toward the TypeScript/React design system over time.

On the GTM and sales side, publicly detectable signals include Salesforce (CRM, confirmed on Himalayas and multiple tech-stack aggregators), Zapier for workflow automation, Zendesk for customer support operations, and Google Workspace (Gmail) for internal communication and productivity. The StackShare listing also shows Shopify, which likely covers merchandise or peripheral e-commerce workflows rather than any core product. Zoom and Webex both appear in job listing requirements and internal tooling references, reflecting the remote-first, globally distributed workforce. These signals are directional; Kraken's institutional sales team likely uses additional sales engagement and marketing automation tools not yet publicly detectable.

Kraken's Rust-first, Kubernetes-native, AWS-and-on-prem hybrid architecture signals a highly sophisticated engineering culture that evaluates vendors on performance, reliability, and security above cost. The active migration from PHP to Rust means legacy PHP-based integrations are in active wind-down — new tooling should natively support Rust, gRPC, and Kafka-based streaming to be credible to Kraken's engineering team. Cloud cost management tools that handle hybrid AWS-plus-on-premises environments (FinOps platforms with on-prem visibility) are a natural fit as Kraken scales post-NinjaTrader.

The IPO preparation creates a concentrated near-term window for vendors in financial close and reporting, SOX-readiness automation, audit tooling, and investor-relations platforms. Kraken's compliance burden under MiCA (EU, 30 EEA countries) and the Wyoming SPDI and Fed master account (U.S.) makes identity verification, transaction monitoring, and regulatory reporting software high-priority procurement categories — and Ben Gray's CLO function controls these vendor decisions. These are not nice-to-have purchases but regulatory requirements with deadlines tied to the IPO timeline.

On the displacement and integration side, the NinjaTrader acquisition is likely generating demand for unified data infrastructure that bridges Kraken's crypto exchange market data (Kafka-based, Rust-native) with NinjaTrader's traditional futures data (CME, CBOE, NFA-regulated). Vendors offering market data normalization across crypto and traditional asset classes, or risk management systems that handle both, will find a genuinely receptive audience at the combined Kraken-NinjaTrader engineering organization. The February 2026 acquisition of Magna (token management platform) creates similar integration demand in the tokenized-equities space, particularly for custody and on-chain settlement tooling that bridges TradFi equity infrastructure with the ERC-20 and Solana-based xStocks architecture.