Product Engineer; m​/f​/d

At Stackgini we are turning business ideas into technology impact.

In most large companies this journey is broken. A good idea lands in an IT organization that runs hundreds of applications, and nobody can tell you quickly what already exists, what is redundant or what to buy. So decisions happen in spreadsheets and gut feeling, and ideas wait.

We are changing that. Our platform understands the enterprise application landscape bottom-up, from individual demands to full portfolio interdependencies, and reasons about it the way the best human experts do, just at machine scale. Our vision: enterprise IT that is self-governing and self-optimizing. No longer a bottleneck, but the engine of innovation.

We are 11 people, and we build this with Siemens Healthineers, REWE, SAP, DATEV, Danaher and other leading enterprises including even more DAX 40 and Fortune 500 companies. We have thousands of users and process very large IT landscapes, one of our customers even exceeds 11,000 applications. At this size there is no hiding. What you ship goes to customers fast, and the people who join now set the standard for the product, the culture and the category.

The role

A Product Engineer at Stackgini owns features end to end. You take a customer problem, scope it with Product & CSM, design the solution, build it across the stack, ship it, and stay close to it after release. Recent examples:
Product Engineers shipped Decisions and Approvals (a feature for routing IT requests through multi-step approval workflows), built evaluation infrastructure to benchmark agent recommendations against expert baselines, and agentic IT solutions research with our inhouse developed agent tool‑use.

You are close to users. You join customer calls, run user tests, read support context, and you use that input to decide what should exist and how it should behave. The person building the feature is expected to shape the feature.

We have more good ideas than hands. For most reasonable feature ideas the question is not whether we build them, but when. Product Engineers are the people who turn that prioritization into shipped product quickly, without lowering the bar.

The AI work

AI is core to our product, but we don't run agents for everything. We use them where they create real value. For that we built our own LLM harness from hard‑earned production experience: maximum freedom for the model, without compromising on enterprise control, governance or EU data residency.

What this means for you: you work hands‑on with these new technologies, you run evals instead of trusting demos, and you know how to put this stuff into production reliably. Agent output gets benchmarked against expert‑curated baselines, and failure modes get named, measured and fixed. If you are obsessed with how AI systems behave, fail and improve, this is the place for you.

How we ship

• Small teams of 2 to 3 engineers around a topic. We deploy continuously to production.
• We use AI tools heavily, and you have freedom in which ones (e.g. Cursor, Claude, Codex). Around that we run a strong operating system: clear ticket context, focused branches, AI review before human review, screenshots or videos for product changes, manual product testing before release, production verification after.
• The engineer who ships a change owns it in production. If something breaks, you lead the fix.

What we need from you

Nobody at 11 people knows everything. We hire people who dive in and own outcomes.

• 3 to 6 years of full‑stack experience shipping production software people actually use.
• You can get productive in our stack fast:
  Kotlin/Spring Boot and Python on the backend, Type Script/React on the frontend. You don't need to know all of it on day one, but you should get into it quickly.
• You have shipped agentic systems, side projects absolutely count. You think in tools, context and evals, not just prompts.
• Fluent English. German is a real plus because many customer calls run in German. We pair you with a German‑speaking teammate when needed, but German gets you closer to users faster.

You'll stand out if

• You build with AI natively, not as an afterthought.
• You can take a vague problem to a…