Applied Mathematician

About the Role

Join Circonomit, a company pioneering decision infrastructure by enabling industrial organizations to tackle complex combinatorial problems through advanced n-dimensional optimization. Founded by researchers from RWTH and supported by a significant €2.8M funding round, we empower German Mittelstand and European industries to make data-driven decisions amidst shifting market dynamics involving orders, machines, and personnel.

We operate with agility, a commitment to problem-solving, and a culture that demands solutions without delay.

Your Mission

As an applied mathematician on our team, you will shoulder responsibility for the mathematical core of our decision engine. This engine translates production constraints, resource capacities, costs, and other parameters into optimized actionable plans for industrial customers. It bridges rigorous mathematical modeling with practical usability for non-mathematicians, ensuring accuracy and accessibility.

You will define the mathematical abstractions and enhance the engine's capabilities, while collaborating closely with engineers and end-users to model real-world planning scenarios, learning from data challenges and pushing the solver's boundaries.

Key Responsibilities

• Develop and own the mathematical foundations behind the models, ensuring consistent and precise abstractions for optimization solvers.
• Convert complex planning problems—capturing capacities, costs, lead times, and shift schedules—into robust models that guide plant managers’ decisions, including preprocessing and validating often messy ERP and Excel data sources.
• Create interpretable and actionable outputs for users, enabling planners to understand improvements, justify plans in meetings, and identify constraint conflicts when ideal solutions are unattainable.
• Scale the solution horizontally and vertically to support multiple sites, longer time horizons, tighter constraints, and concurrent solving for multiple customers without performance loss.
• Manage end-to-end ownership over features, from concept through production deployment and ongoing operation, with direct responsibility for quality and delivery.

Work Culture

We embrace a small, flat team structure with rapid communication and no hierarchies. Each team member is accountable for their projects from development to production. We foster open two-way feedback through daily cooperation and regular one-on-one sessions. Peer code reviews and frequent in-person collaboration in the Cologne office drive fast, insightful discussions, particularly between mathematical modeling and engineering aspects.

Candidate Requirements

• Strong expertise in applied mathematics, especially discrete optimization and algorithm development, with the ability to design modeling abstractions usable by other engineers. A PhD or equivalent practical experience is essential.
• Experience delivering industrial optimization solutions (MILP, CP, or both) that remain robust against real-world messy data, strict deadlines, and end-user environments.
• Awareness of solver capabilities and limitations, including CP-SAT and Gurobi, with knowledge of techniques like warm starts, rolling horizon approaches, relax-and-fix heuristics, matheuristics, and pragmatic heuristics when exact solving is impractical.
• Experience operating optimization workloads in production contexts, handling cancellation, timeouts, and parallel execution effectively.
• Proficiency in production-quality Python, including testing, type annotations, code reviews, and performance profiling, especially for numerical code integrating mathematical models and solvers.
• A keen interest in comprehending customer problems and datasets beyond mere mathematical models.
• Collaborative experience working within teams rather than independently.
• Proficiency in German at a C1 level or higher combined with fluent English. Team communication primarily occurs in German while code and documentation are maintained in English.
• Based in North Rhine-Westphalia (Cologne office) or alternatively Munich, Stuttgart, Berlin areas, or willing to work with a hybrid setup. Open to discussing arrangements if there is a good fit.

Preferred Qualifications

• Familiarity with sparse or tensor numerical methods at scale.
• Experience with compiler development, domain-specific languages, or type system design.
• Work on numerical or compiled code performance optimization.
• Understanding of solver internals and deployment of scalable solver workloads.
• Background in production planning, supply chain management, or logistics.

Personal Traits

We seek individuals who are structured, initiative-driven, and possess a winning mindset fostered through diverse experiences. This role is not suited for those preferring research freedom without product focus, wanting to rewrite core engines repeatedly, working only within restricted abstractions, delegating data responsibilities, or awaiting handed tasks.

What We Offer

• Meaningful impact on how factories operate, working on real industrial datasets within a product actively used by customers who quantify your contribution in euros and provide direct feedback.
• Complete ownership over the core mathematical engine influencing product direction and production behavior.
• Collaborative environment with operations research engineers and leadership engagement.
• Fast and transparent feedback on performance and expectations with candid communication.
• Competitive salary paired with equity opportunities (VSOP) reflecting your input and growth trajectory.
• Choice hardware setup to support your workflows.
• Budget allocation for AI tools utilization.
• Sports club membership.
• Access to Deutschland-Ticket for public transportation.

Recruitment Process

The hiring process includes a brief initial call, an in-depth technical interview focused on your modeling and engineering skills, followed by a practical three-hour challenge with a debrief discussion. Final stages involve meeting the team both online and onsite. The total process spans two to three weeks with prompt feedback after each phase. Candidates are encouraged to request candid conversations with current employees to understand the role’s realities before committing.

Application Instructions

No cover letter is needed. Share a production model you developed and highlight one modeling decision you would reconsider in hindsight.