StaffSoftwareEngineer

WE UNLOCK GROUND TRUTH FOR THE ENTERPRISES THAT POWER THE GLOBAL ECONOMY. Billions of critical conversations happen every day in retail, hospitality, food service, logistics, and beyond — with customers and employees. Historically, these insights disappear the moment they're spoken. Arbor turns in-person conversations into executive-grade strategic intelligence. We're signing real enterprise contracts in industries that power the global economy. The opportunity is massive, the problems are unsolved, and we need product-obsessed builders to join on our journey. Must-haves - 5+ years shipping production software, including time as an early engineer at a seed/Series A startup - Full-stack generalist who owns features end-to-end with sharp product judgment — knows what to build, what to cut, when to push back - Ships predictably and communicates clearly; credible with an enterprise VP of Ops on a customer call and async with a remote CTO - Production experience in at least one of voice/audio, LLM apps, or data/insights pipelines (Umi is all three) - Fluent in Python/FastAPI + TypeScript/React; moves across frontend, backend, and infra in a single PR - Has shipped against flaky third-party APIs under latency/reliability pressure (telephony, real-time audio, payments, etc.) - Working knowledge of LLM app development: prompts, structured output, retrieval, evals, and non-deterministic systems that hold up in prod - AWS or GCP with multi-tenant SaaS security instincts (data isolation, PII, SOC 2) - Excited to be a foundational hire building AI for frontline enterprise   Nice-to-haves - Staff/tech-lead IC ready to grow into eng lead within ~12 months - Voice agent stack (LiveKit, Pipecat, Vapi, Retell, or custom on Twilio/Deepgram/ElevenLabs) + speech fundamentals (ASR, TTS, diarization, VAD) - Production agentic LLM workflows: tool use, multi-step, eval-driven - Data/analytics depth: warehouse modeling, DBT, owning an insights pipeline - On-call comfort on a small team — ha