Our Opportunity

Over the past three years, Predium has built the leading real estate intelligence platform for ESG and value management in the DACH region - with the strongest names in the industry as customers: asset managers, housing companies, banks. Our customer base, our data, our credibility - that's valuable capital for the next step. Investment, credit, and asset management decisions in real estate are still made in Excel, case by case. We're now building the AI analyst that takes this work off people's hands - working directly on the customer's own data, scalable without adding headcount. We're well-funded, and we're bringing AI into real estate.

Your Mission

As a Data Scientist at Predium, you will help develop and improve the data-driven models behind our platform. You will use statistics, machine learning, and numerical methods to solve well-defined problems related to buildings, energy performance, and real-estate decarbonization.
Working closely with our engineering and product team, you will turn real-world data into reliable analytical, simulation, and machine-learning models that identify optimization opportunities for more sustainable buildings and help customers make informed investment decisions.

What you'll own

Develop and improve analytical, machine-learning, and simulation models to address business problems in the real-estate and energy domain.

Build and improve data extraction pipelines for structured and unstructured sources, evaluating extraction accuracy and reliability

Continuously explore and evaluate new data sources to improve model accuracy.

Assess data quality, identify gaps and evaluate model performance by validating assumptions and results against real-world data and feedback of domain experts.

Collaborate with data scientists, engineers, product managers, and other stakeholders to translate business needs into data-driven solutions.

What success looks like

You take ownership of well-defined analytical and modeling problems and independently deliver robust solutions.

Your models and analyses are reproducible, well-documented, and supported by appropriate validation.

You translate technical results into clear implications for the product, our customers, and the business.

You collaborate effectively across Data, Engineering and Product teams.

You actively seek feedback and use it to improve both your technical work and your way of working.

What it takes

A degree in data science, computer science, mathematics or a related field.

1–2 years of relevant hands-on experience, with programming, ideally in Python, through projects in areas such as physics-based modeling, machine learning, or data pipelines. TypeScript experience for product prototyping is a plus.

A structured, independent approach to problem-solving.

Curiosity about business needs and the ability to explain findings clearly to technical and non-technical stakeholders.

Confident communication in English; German is a plus.

Bonus:

Initial experience with optimization methods, LLM-based data extraction, or uncertainty and sensitivity analysis.

Familiarity with real estate, building physics, or energy systems.

Experience with tools for physical simulation or energy modeling.

Dealbreakers

Why Predium?

• Stage: ~40 people, well-funded, on the way to defining the category. Early enough that a lot still needs to be built. Yes, it's messy sometimes - that's the whole point. If you're looking for a job with a finished playbook, this isn't the right fit.
  
  
• Impact: You're not cog #100 in some company. The models you build feed straight into our product and get used by real customers. Your contribution shows up directly, in revenue and in the product.
  
  
• Ownership: We don't manage your time, we measure your results. You get responsibility and freedom - and the expectation that you'll use both.
  
  
• Compensation: Competitive salary, bonus, and equity. We want you to share in the success you help create.
  
  
• Team: Office-first in Munich's Werksviertel, 4 days/week.

30 vacation days (including company shutdown days) · Wellpass · Deutschland-Ticket · Jobrad · team events