Senior Product Engineer (Python/AI/Odoo)

About ilexius

ilexius GmbH is a German software company that digitalises automotive aftersales. Since 2005, we have grown from a trade marketplace into one of Europe's leading providers of aftersales software for dealer groups and importers — among them Jaguar Land Rover, BMW, Aston Martin, Peugeot and Volvo.

Our products run in more than 40 countries for 400+ clients and over 10,000 daily users: electronic vehicle health checks (eVHC), warranty and goodwill claim administration (CAT), electronic job cards, online service booking, and the middleware that connects all of it to dealer management systems.

We are headquartered in Wiesbaden, with teams in Serbia and the UK. Our Belgrade office is a small, senior engineering team — the people who build the products, not a delivery outpost.

About the role

This is a new role. We build software for automotive aftersales, and we are changing how we build it. Our team already works with AI agents against our Odoo codebase every day. We are looking for someone senior to push that much further — and to ship real product while doing it.

We are at the point where AI stops being a demo and starts being a product. Our clients — dealer groups and workshops — sit on years of vehicle, claim, and service data, and they want the software to do more of the thinking. We are looking for someone to lead that work.

You will be hands-on across our roadmap. What makes the role different is the leverage. You are expected to design and prototype ideas quickly enough to put them in front of a retailer within weeks.
Some of what you build will have AI inside it. Most of it will simply be built with AI. Both matter — the second is the bigger opportunity right now.

You will also spend real time with clients, understanding how a service advisor actually works rather than how a specification says they do.

It is a product engineering role, not a research role. We care about what ships.

What you'll own

• Prototypes that reach clients fast — taking an idea from a retailer conversation to something clickable, testing it with them, and being honest about which ones should die. Then hardening the ones that stick into production Odoo modules.
• Building with AI, as the default — we already use Claude Code against our Odoo codebase. You will push this further: agent workflows, MCP servers against our GitLab, Postgres and Notion, and skills that encode how we build. Making the whole team meaningfully faster is worth as much as any single feature.
• Prototypes that reach clients fast — taking an idea from a retailer conversation to something clickable in days rather than months, testing it with them, and being honest about which ones should die. Then hardening the ones that stick.
• The client conversation for your area — running discovery calls and demos with retailers, translating what they say into something buildable, and closing the loop when it ships.
• AI features in our products — LLM-powered extraction from claim paperwork, pre-checks that catch a warranty claim before it is rejected, summarisation of vehicle health checks, natural-language search across service history. You help decide which problems are worth solving, then build them.
• Reference: M6xPZO-xNwQp-TamIdCx8ZuXQXM0tUbj8b7nHrjOVZyXJHK14UG_RGWsts6BeuPN

What we're looking for

• At least 5 years of professional software engineering experience, with Python as your main language
• You already build with AI every day, and you have opinions about it: which tools, where they genuinely help, and where they quietly cost you time. Coding agents, context and prompt design, reviewing what a model gives you rather than trusting it.
• Enough hands-on LLM work to build a feature with one — RAG, tool and function calling, and the judgement to tell whether the output is actually good enough to ship. Substantial side projects count.
• Strong API integration skills, SQL, and relational data modelling
• The instinct to prototype fast, and the discipline to then build it properly: tests, reusable code, review
• You can sit in a call with a non-technical client, ask the right questions, and leave with a clear picture of the real problem
• Working knowledge of Linux, Git, and Docker
• Fluent English, written and spoken
• A degree in computer science, or equivalent practical experience

Nice to have

• Odoo or another ERP. We will teach you Odoo, but knowing it shortens your first months.
• Automotive, aftersales, or dealer-management-system experience
• Frontend skills (JavaScript, Odoo web / OWL) for making prototypes real enough to test
• Basic German — many of our clients are German-speaking
• Public work: GitHub, writing, talks

What success looks like

In your first months, you get deep enough into one of our products to form your own opinion on where AI genuinely helps, and where it is a distraction. Over your first year:

• At least one feature with AI you built is in production and used daily at our clients, and you can point to what it changed for them
• Retailers ask for you by name in discovery calls, because you understand their business and not only our software
• The rest of the team builds differently because of tooling and practices you introduced
• Several of your prototypes have been killed — quickly, cheaply, and with something learned from each

How you'll work

• A small, senior team in our Belgrade office. Hybrid: a few days in the office, flexible hours, home office the rest of the time.
• Direct access to the people who decide what gets built, in Belgrade and in Wiesbaden. No layers in between.
• A mature Odoo codebase with real users, Docker-based development environments, GitLab CI and a genuine test culture. This is a company and product with 20 years of history behind it — you will respect what is there while changing it.
• Regular contact with clients across Germany, the UK and beyond. Occasional travel if you want it.
• Room to be wrong. Prototyping means most ideas do not survive; that is the point.

What we offer

• Above-average salary and a permanent contract
• Private health insurance
• Flexible working hours and a home office policy
• Paid AI tooling — Claude, Codex, API budget for experiments. You should never wait for approval to try something.
• Dedicated time for experiments that are not on the roadmap
• Budget for training, courses, conferences and books
• A fully stocked kitchen (coffee, tea, fruit, soft drinks, snacks)
• Skilled, senior colleagues and products that thousands of people rely on every day