Build the algorithmic core of planqor.ai: the real-time decision layer that steers discrete manufacturing when production plans get disrupted. Born on the shop floor, co-founded by senior production experts, and with direct access to high-impact industry problems. Work directly with our CTO and shape the optimization architecture on a greenfield: custom algorithms, heuristics, OR and simulation, fast C++ with Python on top. Backed by investors who have built and scaled successful companies. Remote-first or from our office in Munich, competitive pay + VSOP, no puzzle interviews.
Factories have invested billions in automation, yet production is still steered largely by people. Every night, a plan is computed for the next day, and the first disruption makes it obsolete. From then on, planners and shift leads juggle dependencies by hand. That is expensive, rarely optimal, and wastes much of what these factories were built for.
We are building the layer that takes over when the plan gets disrupted. It connects to factory data, models production, detects disruptions and bottlenecks as they emerge, simulates responses and recommends the best one.
planqor.ai was born on the shop floor. Our founding team brings deep production expertise, strong industry networks. We have the runway to build deliberately and the experience of people who have built and scaled successful companies before. On the technical side, we combine deep algorithmic and software-engineering competence with modern optimization methods, GenAI-assisted development and fast, pragmatic iteration. We are in stealth and very early, which means you can shape much of what comes next.
The algorithmic core: Design how planqor.ai represents, propagates and optimizes schedules, resources and constraints, and how it repairs them when disruptions hit.
Performance: Make it fast. Profile and optimize the C++ core and expose it cleanly.
Evidence: Build the benchmarks, instances and experiments that show what actually works and where it fails.
The path to production: Work with our systems engineer to turn algorithms into decisions planners trust.
The team: Grow into leading optimization at planqor.ai as we scale: hiring, mentoring, setting direction.
You set your own priorities and structure your work toward clear goals. No ticket queue, no micromanagement.
Our goal is not to build the most impressive optimization algorithm on paper. It is to build decision systems that remain useful when the real world changes. A strong solution that can be updated, repaired and explained quickly beats closing the last optimality gap on a static model.
So we are neither a solver wrapper nor a research lab. We use whatever works (MIP, CP, heuristics, local search, simulation, learning or something purpose-built) and let the evidence decide.
Generative AI makes this loop dramatically faster. We use it to implement ideas, explore alternatives and evolve algorithms against our benchmarks. But speed only counts if the methodology holds: representative data, adversarial tests, meaningful metrics, rigorous validation.
There are no points for cleverness. The simplest system that reliably solves the real problem wins.
Strong modern C++ and performance engineering. You are comfortable building performance-critical algorithms, profiling them, choosing appropriate data structures, and reasoning about runtime, memory, and scalability.
Real experience solving optimization problems. This may come from Operations Research, combinatorial optimization, scheduling, MIP/CP, heuristics, local search, simulation, optimization under uncertainty, or related fields. You do not need to be an expert in all of them. Big plus for Sequential Decision Analytics.
Systems thinking. You think beyond individual models and algorithms: how state is represented, how changes propagate, how components interact, and how the complete decision system behaves under real-world disruptions.
Python for experimentation and data analysis. You can build benchmark pipelines, analyze data, visualize results, and quickly investigate why an approach succeeds or fails.
Empirical engineering judgment. You design experiments, challenge assumptions, measure performance, and abandon approaches when the evidence says they are wrong.
Independence. You can turn an underspecified operational problem into working software without waiting for a detailed spec.
Fluent English. German and experience in manufacturing or industrial planning are useful but not required.
Ownership: shape the optimization architecture from line one, working directly with our CTO.
A clear growth path into leading the optimization team.
A real problem worth solving.
We equip you with the hardware, compute and tools you work best with.
Remote-first or from our Munich office, with regular in-person workshops.
Competitive pay + VSOP.
Send your CV.
Talk with our CTO about your work and how you approach hard problems.
Meet our CEO and selected investors.
More at home in backend and product? Check out our Founding Full-Stack Engineer role.