Principal Software Engineer, Backend

Yes is an AI native online casino and sportsbook preparing to launch under a Danish Gambling Authority (DGA) licence. We run on the EveryMatrix platform for wallet, payments, casino, sportsbook and bonuses, and build our own services on Google Cloud (GKE and Cloud Run) to own the player experience across iOS, Android, mobile web and desktop. We are a small senior team. Every engineer ships, owns what they ship, and works in a regulated environment where availability, auditability and player protection are not optional.

About the role

Our backend layer is a set of Backend-for-Frontend (BFF) services, written in Java, that sit between our clients and the vendor platform. They aggregate, shape and cache data for each client, handle both request-response and real-time event traffic, and enforce security and regulatory rules before anything reaches the player. As a Principal Backend Engineer you set the reference architecture for these services and are accountable for how fast and how reliably they run. You will write production code, make the hard design calls, and raise the bar for every backend engineer who joins after you.

What you will own

• Scope BFF services and their APIs to all clients.
• Integration with EveryMatrix and other vendor APIs and event streams.
• Caching, data stores and messaging used by our backend.
• Backend observability, SLOs, load and capacity testing.

Must have

• Seniority: 10+ years building backend systems, with at least 3 in a lead, staff or principal role setting architecture across multiple services.
• Java at depth: modern Java (17+), JVM internals, memory and garbage-collection tuning, concurrency (including virtual threads and reactive or non-blocking I/O), and profiling under load.
• Design patterns and architecture: practical command of GoF and enterprise integration patterns, hexagonal or clean architecture, domain-driven design, and API gateway / BFF patterns, plus the judgement to know when not to use them.
• Spring Boot (or equivalent framework) in production at scale.
• Sync and async communication: REST and gRPC; event streaming and messaging (Kafka, Pub/Sub or similar); real-time push to clients (WebSockets or server-sent events); delivery guarantees, ordering and idempotency.
• Caching: distributed caches (Redis or similar), multi-level caching, invalidation strategies, and Data: relational databases (PostgreSQL or similar) including query and index tuning, and the consistency trade-offs of distributed data.
• Performance engineering: load and soak testing, latency analysis at p95/p99, capacity planning, and evidence of having made a production system measurably faster.
• Availability and observability: SLOs and error budgets, metrics, tracing and logging (OpenTelemetry, Prometheus or similar), and hands-on incident leadership.
• Cloud-native delivery: containers, Kubernetes, CI/CD and safe deployment practices (canary, feature flags, rollback).
• Security: OAuth2/OIDC, token handling, input validation and secure-by-default API design.

Strongly Preferred

• Other backend languages: Kotlin, Go or Node.js. We choose the BFF language per service, so breadth matters.
• Regulated or high-traffic real-time domains: iGaming, sports betting, payments, trading or similar.
• Experience with EveryMatrix or comparable iGaming platforms (wallet, cashier, casino aggregation, sportsbook feeds).
• Google Cloud (GKE, Cloud Run, Memorystore, Pub/Sub, Cloud SQL), Cloudflare, ArgoCD and Terraform.
• Server-driven UI or aggregation APIs serving several client platforms.
• Compliance-driven engineering: DGA SCP, PCI DSS, ISO 27001 or GDPR.

What success looks like

• First 30 days: Understand the current BFF services, vendor integrations and traffic profile. Establish a performance and availability baseline and publish the gaps. Ship production code in at least one BFF service.
• First 90 days: Publish the BFF reference architecture covering service structure, sync/async patterns, caching and data rules. Agree SLOs for player-critical journeys (login, deposit, bet placement, game launch) with dashboards and alerts behind them. Run the first full load test against launch traffic forecasts and close the critical findings. BFF layer meets agreed SLOs under forecast peak load, with tested degradation paths for every vendor dependency. Runbooks and on-call in place; backend engineers designing new services to the reference architecture without your involvement.

Compensation

Competitive base salary. Full range and benefits shared at first conversation. Intro call (30 min): role, team and your background. Technical interview (2 hrs): a deep dive on Java, the JVM, concurrency and design patterns, using a system you built; system design of a BFF for a real-time sportsbook or casino journey covering sync and async flows, caching, data and SLOs; and a pair-programming exercise on a realistic backend problem. Leadership and values with the CTO and HR (45 min).