Zp-Inf-Infrastructure Engineer · model Routing Omaha​/Remote · HYBRID Platform

Build the layer that decides which model answers which request, at what latency, for what cost. Runs at the critical path of every customer integration.

• Design and operate the routing layer that dispatches requests across Claude, GPT, Gemini, Grok, and customer fine-tunes.
• Own latency and cost as first-class observability signals -- not afterthoughts.
• Build the fallback and retry logic that keeps integrations alive when a provider has an incident.
• Instrument everything: request traces, model call logs, cost attribution per customer.
• Work directly with forward-deployed engineers to meet per-engagement SLAs.

• Deep experience with distributed systems under real load -- not demo load.
• Have built or operated a service with strict latency SLAs (p95 < 100ms class).
• Understand LLM provider APIs well enough to know where the failure modes live.
• Strong instinct for when to add abstractions and when to just write the code.
• Comfortable on call for a system that other engineers depend on to ship.