Backend/Systems Engineer, webAI

• 3–6 years of backend/systems engineering experience (or equivalent impact)
• Strong fundamentals in distributed systems: consistency models, replication, idempotency, failure modes, and performance tuning
• Experience building production services in at least one systems-friendly language (e.g., Go, Rust, Java, C++, Python) and comfort reading others
• Familiarity with real-time systems: WebSockets, gRPC streaming, pub/sub, queues, or event-driven architectures
• Experience designing APIs and data models with security and privacy in mind
• Working knowledge of cloud infrastructure (AWS/GCP/Azure), containerization, and CI/CD

Desirable

• You’ll build the “small but serious” backend systems that power real-time, privacy-preserving product experiences like identity, sync, encrypted state, messaging/presence, and group conversations without defaulting to a giant, monolithic backend
• You’ll think in terms of edge + device + cloud as a single system, and you’ll ship pragmatic, scalable coordination services that keep products fast, reliable, and secure
• Implement auth flows, device registration, key management integrations, token/session lifecycle, and account recovery patterns
• Support multi-device identity and secure device-to-account bindings
• Build APIs and services that enable state sync across devices (event-based or state-based replication)
• Partner with client engineers to support end-to-end encrypted (E2EE) state where the server coordinates without needing plaintext access
• Create “coordination services,” not “big backend”: rendezvous, discovery, sequencing, fan-out, rate-limits, abuse controls, and reliability tooling
• Prefer minimal durable state and clear boundaries; push logic to the edge/client when appropriate
• Implement reliable messaging pipelines (delivery semantics, retries, ordering strategies, idempotency)
• Support presence and group membership (joins/leaves, device fan-out, read receipts/typing signals as needed)
• Integrate with pub/sub or streaming systems, and tune for latency
• Design for intermittent connectivity, offline-first behavior, device constraints, and high tail-latency networks
• Make trade-offs explicit: consistency vs availability, ordering vs throughput, durability vs cost
• Add observability (metrics, tracing, structured logging), define SLOs, and participate in on-call/incident response (lightweight rotation)
• Improve reliability: backpressure, load shedding, graceful degradation, and safe rollouts