# Senior Product Engineer (AI, Full-Stack) at Azx

- Company: Azx
- What the company does: Tech transformation, societal shifts, and environmental disruption put enormous pressure on critical industries like energy, infrastructure, real estate, and others to adapt, pursue new growth opportunities, and optimize operations. Backed by Powerhouse Ventures.
- Company website: https://azx.io
- Type: Startups (AI role)
- Level: Senior
- Location: Seattle
- Work setup: Remote
- Pay: $140K to $230K base salary per year (USD)
- Posted: 2026-08-31
- Apply by: 2026-10-15
- Apply: https://jobs.ashbyhq.com/careers.azx.io/ab44ab2b-0df3-4492-91bf-ccc90099a56e
- Page: https://www.1752.vc/careers/jobs/azx-senior-product-engineer-ai-full-stack/

## About the role

We are seeking a Product Engineer that owns the demo's we build for clients from the schema to the pixel, the deployment and the observability. You'll work across the chat surface and agent workspaces, the app builder alongside the platform team, and the shared component and design system, plus the agent-facing developer experience — since our platform's users are coding agents as much as humans.

## What they're looking for

- 5+ years of deep, production experience with React/TypeScript — streaming UIs, high-frequency updates, virtualization, and real state management, with informed, hands-on opinions about React 19
- Hands-on experience with a streaming protocol end to end (SSE or WebSocket) — chunk boundaries, reconnection semantics, abort propagation — including having personally debugged a raw byte stream
- True full-stack ownership : you've shipped the backend half of your own front ends (Node/Fastify or Python/FastAPI), handling auth, idempotency, pagination, race conditions, and migrations
- Product taste backed by enterprise experience — accessibility you can defend to a screen-reader user, thoughtful empty/error states, and localization that goes beyond a strings file
- An AI-native working style: daily use of AI coding tools, driven by specs rather than vibes, with concrete examples of where you overrode a tool's output
- A strong sense of ownership — instrumentation, evals on AI-dependent behavior, cost awareness, and on-call are part of "done," not a later phase

Tags: Engineering