Agentic Engineer (f/m/d)

We are a startup digital agency, and no two projects, clients, or even days look the same. But all of our days are agentic-first.

In ecommerce we build on Shopify and Liquid and Next.js, with custom AWS microservices and apps that take Shopify well past the defaults.

On the marketing side we build highly customised sites in Next.js and React, with careful design and animation and a real focus on how the page performs and feels.

In automotive and other industries we build full platforms, AI tools and integrations, from stock-car locators to configurators and a lot more.

Across all of it, we bring AI into the work through our own products.

Here nothing will be fixed. You will work across changing projects and settings, on long builds that run for months and shorter spikes to figure out how a business can actually use AI, and you will help us rethink how agency work is done.

Sound like your kind of place? Here is what we need from you.

We're looking for engineers who are excited by agentic and AI-native systems - turning models into products that hold up under real users, real load, and real time. You'll build LLM integrations, agents, and RAG pipelines, treat prompt engineering as a code discipline, and make systems reliable through evaluation, guardrails, and solid testing. Whether you're building your craft or already shaping the AI architecture of a project, we want to hear from you. We hire across levels - here's what each looks like at OH-SO:

Junior AI/Agentic Engineer

You're building your engineering craft - curious about why and how things work, not afraid to get your hands dirty, and eager to grow. You implement from clear briefs - LLM integrations, prompt scaffolding, RAG components - and write the tests and evaluations that keep them honest, comfortable using AI coding tools (Claude Code, Cursor, Copilot) under supervision. You want to understand why and how we're building things, ask what the purpose and the constraint are before you start, and your work is reviewed before it merges.

AI/Agentic Engineer (typically 2-3+ years)

You own an AI feature end-to-end - an agent flow, a RAG pipeline, an integration across its error paths and guardrails - delivered independently with light oversight. You build evaluation harnesses to prove quality rather than assume it, use AI coding tools autonomously in production, and contribute to the shared AI backbone (the AI-brain MCP and skill library).

Senior AI/Agentic Engineer (typically 5+ years)

You own the AI outcome of a project - you define the integration shape, the multi-agent architecture and orchestration, the evaluation strategy and guardrails, and you write the ADRs that document why. You're the technical credibility in the room, contribute patterns and skills back to the shared backbone, and mentor the engineers around you. You use AI to define new ways of working, and you think in long runtime: the AI landscape will shift underneath the project, so you build for inheritance and replaceable layers. You help us move forward as an organization and you act as an AI evangelist.