Staff Software Engineer - Distributed Systems & Edge Platform

**Omara builds operational software for governments and highly regulated institutions.** We are hiring a Staff Engineer to build our **distributed systems and edge platform**: offline-first applications, data replication and synchronization, resilient local infrastructure, device fleet management, and on-premise AI inference. Our systems operate across unreliable networks where devices may remain disconnected for days and still need to create, modify, and serve critical state. We build for the last mile: software that must remain useful in the field, not just inside a cloud region. #### About the Role We are looking for a Staff Engineer to own major parts of Omara's distributed edge platform. Our applications run across edge appliances and user devices that may operate without connectivity for hours or days. Devices create and modify state independently, software versions drift, hardware fails, and local applications must continue operating. When connectivity returns, the system has to reconcile safely without losing, duplicating or silently corrupting state. You will make the architectural decisions that determine how this works. This is a hands-on individual contributor role. You will design systems, write production code, debug failures in the field, and set technical direction across the platform. #### What You'll Work On **Distributed state and synchronization** - Design offline-first replication and reconciliation across devices and sites - Define identity, idempotency, conflict and deletion semantics - Build durable local state, mutation logs, checkpoints and resumable synchronization - Handle partial failure, device loss and long-lived network partitions - Evolve protocols and schemas across mixed-version fleets **Edge platform** - Build the runtime for deploying and operating applications on distributed appliances - Own provisioning, device identity, networking, observability and remote diagnostics - Design safe software upgrades, rollbacks and recovery - Operate containerized workloads using Kubernetes/K3s or similar systems **Local compute** - Operate GPU-backed inference alongside conventional application workloads - Deploy and manage runtimes such as SGLang, vLLM or equivalent - Make local model serving reliable, observable and easy for applications to consume AI is one workload on the platform. This is fundamentally a distributed systems role. #### What We're Looking For You have 10+ years building production systems and have personally owned consequential architectural decisions. We are particularly interested in engineers with experience in one or more of: - database replication or synchronization - distributed storage or databases - offline/local-first systems - edge computing or infrastructure appliances - device or server fleet management - networking or systems infrastructure You should have strong fundamentals in distributed systems, operating systems, networking and databases, and be comfortable reasoning about consistency, concurrency, idempotency, failure recovery and protocol evolution. Strong systems programming experience is required. #### The Kind of Problem You'll Own Eight devices operate independently for three days. All eight can create and modify records. Two are then permanently lost. The remaining six reconnect with overlapping and conflicting state. What should converge automatically? What should never be merged? What guarantees are no longer possible? How does the answer survive years of schema and software evolution? If that is the kind of problem you enjoy, we'd like to talk. #### Why Omara We are a small team working on systems that have to survive the physical and institutional realities of government: unreliable connectivity, fragmented infrastructure, sensitive data and users operating far from headquarters. There is little distance between architecture, implementation and production. We expect Staff Engineers to help determine not only how systems are built, **but what should be built at all.** We already have paying public-sector customers and live deployments.