Agent Experience Engineer

Enterprise teams still copy data between systems all day. Work gets stuck in emails, legacy UIs, and handoffs. That chaos is costly, slow, and risky.

We're a fast-moving team on a mission to end it for good. Traction is strong and we're solving real problems for real customers—but to win, we need exceptional talent. We stay humble, do the work, and let results speak.

We're building the AI operations platform for retail and CPG enterprises—a horizontal platform where AI agents execute end-to-end work across UIs and APIs with governance built in.

Where copilots stop, Duvo finishes the job. Business users specify the outcome; agents plan, act, request approvals on exceptions, and learn with every run. We start with a retail wedge (category management, supply chain, finance ops) where ROI is obvious, then expand to adjacent functions and sectors.

Velocity is our moat: ship fast, iterate faster, compound learning.

Think Developer Experience, but for agents — you own the experience of the agents that build and run Duvo.

As an Agent Experience Engineer (AX), your job is engineering-team efficiency. You make our agents fast, correctly wired, and easy to run end-to-end. The speed of the build/run loop is the product: when agents can iterate faster, Duvo moves faster.

This role owns the plumbing that everything else runs on: CI/CD, the monorepo, agent tooling, and the environments that let agents drive the real app. You'll also build agents and internal workflow apps around deployments, runs, and CI/CD — the meta-layer that helps engineers and agents stay in flow.

These are non-negotiables—the things we'll specifically evaluate you on:

• CI/CD mastery. You've built and maintained fast, reliable build/run pipelines. You understand where time is lost and how to get it back.

• Tooling and infrastructure. You own the monorepo, build systems, and developer tooling. You think about ergonomics for both humans and agents.

• Self-hostable environments. You can set up cloud desktop or browser-in-browser environments so agents can drive the real web app — enabling strong end-to-end tests without brittle mocks.

• Agent wiring. You understand what it takes to connect an agent to the right resources: internal sites, tools, APIs, actions, and pipelines. You wire things correctly and maintain that wiring as the platform evolves.

• Tight feedback loops. You care deeply about making things runnable from anywhere — local, CI, staged — with fast, useful feedback.

• Building agents and workflow apps. You build internal tooling — agents and apps — around the deployment, run, and CI/CD surface. The meta-layer isn't an afterthought; it's part of what you own.

• Have experience with SRE, platform engineering, or DevEx.

• Have worked on systems where agents or automation (not just humans) are first-class users of your tooling.

• Care about how data and internal systems are accessed — you'll also shape how agents access and act on internal data and entities.

• Have worked with monorepo tooling (Nx, Turborepo, Bazel, etc.).

• Have experience with cloud desktop or headless browser environments.

• You want to build product features end-to-end — see Product Engineer or Tech Lead.

• You're not interested in going deep on tooling, infrastructure, and developer experience.

• TypeScript-first

• React and Fastify

• Postgres, GCP

• Latest AI primitives

• Initiative-driven. Organize around customer problems, not org charts.

• Customer-obsessed. Features that don't move customer metrics get cut.

• Iterative by default. Ship small, learn fast, fix forward.

• AI-first leverage. If a tool can do it, a person shouldn't.

• Direct feedback. Actionable feedback, given immediately.

• Autonomy with accountability. Outcomes over process.

• Unlimited AI budget.

• Autonomy to do your best work.

• A real AI product with real customers.

• A sharp, motivated team that values ownership and candor.

• Competitive compensation with a meaningful equity component.

1. Hiring manager screen (30 min).

2. Remote task (async, time-boxed, ~1 hour).

3. Technical interview (Prague, ~1 hour).

4. On-site trial day (2 days), fully compensated.