Open a news app during German automotive earnings season and you will find some version of the same headline: restructuring, cost pressure, another plan to cut. If you are a student weighing a future in or near this industry, the natural reaction is to read that as a sign to stay away. That reaction misses what is actually happening underneath the headlines.
What is actually shifting
Three structural changes are moving through German automotive at once. Electrification means fewer of the traditional mechanical components that used to define a combustion engine, and different assembly and diagnostic skills around the ones that replace them. Software-defined vehicles means a growing share of what makes a car good now lives in its code rather than only in its mechanics. Cost pressure across the industry is pushing companies to automate and digitise processes that used to run on paper and manual checks.
None of that is a claim that the industry is shrinking or growing. It is a claim about which skills the industry is buying, and that mix is changing faster than the headlines suggest.
The skills the industry is buying more of
Less pure mechanical routine work. More software engineering, more data analysis, more quality analytics, reading sensor and inspection data at scale to catch problems earlier than a manual check ever could, and more supply chain capability, the same overlap of planning, procurement, logistics and data work covered in a lot of automotive reporting but rarely explained in plain terms.
This is not an argument that mechanical skill stops mattering. It is an argument that the mechanical engineer who can also read data, or write a script, or reason about software, is becoming noticeably more employable than the one who cannot.
Suppliers matter as much as the famous brands
The big vehicle brands get almost all of the headlines, good and bad. A large share of the actual engineering, and a large share of the actual hiring, happens one or two tiers below them, at the companies making the parts, the sensors, and the embedded software that goes inside those vehicles. Those companies rarely make the news, which means students searching by brand name alone are missing a big part of the market.
Where legacy software modernisation fits in
A large amount of the code running inside automotive and industrial companies today is old. It was written years ago by people who have since moved on, and it now needs to be understood, maintained, and carefully modernised rather than thrown out and rebuilt from scratch. Reading someone else's old code, working out what it actually does, and safely bringing it forward is a real, hireable skill, and increasingly one done with an AI coding agent doing a share of the heavy lifting.
That is exactly the space an exciting IBM Bobathon challenge, built with IBM Bob, is aimed at. IBM Bob is an agentic AI software development tool: it plans, writes, tests and modernises code across a whole repository, not a data science tool. The Legacy Fix scenario inside the challenge puts you in your first week as a junior engineer at a constructed German car-parts company, working through a legacy Python codebase that needs fixing and modernising. It is a training scenario built for the challenge, not a real company's real fault data, and any AI coding agent can be used to work through it, with a free trial available for IBM Bob itself if you want to try the tool the challenge is built around.
Finish the challenge and you can earn an IBM SkillsBuild certificate of completion, free to take. It is not automatic: about three in ten participants do not pass, based on ProoV's own grading data, so it carries real weight on a CV precisely because it is not guaranteed.
The overall picture is less dramatic than the headlines and more useful to act on: build proof in the areas where mechanical and software skill overlap, and do not limit your search to the handful of brand names everyone already knows. For a wider view of the German job market heading into next year, see our 2026 outlook, and for the full detail on the Bobathon challenge itself, visit the IBM Bobathon page.
