Senior Backend Engineer (f/m/d)

Our Opportunity

For the last three years, Predium has built the leading ESG intelligence platform for institutional real estate companies in the DACH region - with the biggest names in the industry as clients: asset managers, housing associations, banks. Our customer base, the data, our credibility are valuable assets for the next step. Investment, credit, and asset management decisions in the real estate industry are still made in Excel, case by case. We are now building the AI analyst that will take over this work. Directly on the customer's data, scalable without additional staff. We are well-financed and are bringing AI to the real estate industry.

Your Mission

As a Senior Backend Engineer, you will be responsible for the systems behind our platform: the GraphQL API, the relational data model, the background processing that transforms our customers' documents and Excel files into structured data, and the authorization model that allows a single owner and a bank with hundreds of portfolios to use the same product. The majority of your work will go from schema design to resolver. You will work closely with Product, Design, and the other engineers to build what matters – not just what you've been told. You will take real responsibility for production: If something is slow, you will find out why – whether the answer is a bad query, application code, or the infrastructure underneath. You will also help us make AI reliable in production. Document extraction, long-running analysis jobs, and agentic features are central to the product's direction – and we need people who make these workflows testable and reliable, rather than just impressive in demos. With us, the best idea wins – no matter who it comes from. With 40 people, there's simply no room for politics. And: every line of code contributes to improving the carbon footprint of real estate.

Tech Stack: TypeScript (NestJS in backend, React in frontend), Hasura/GraphQL, PostgreSQL (multiple databases including PostGIS), Queue-based job processing, AWS (ECS / Elastic Beanstalk, S3, SES, Textract), Infrastructure-as-Code, Claude

What You Will Do Specifically

• You will build features across our Hasura GraphQL API, NestJS, and Postgres - from schema design to resolver.
• You will model complex real estate and ESG domains in relational schemas. We operate multiple Postgres databases in parallel (one of them geospatial), and schema decisions have long-term consequences here.
• You will be responsible for background processing: Excel imports, document extraction, long-running analysis jobs. Queues, retries, jobs that start other jobs - and the question of what happens if step 3 of 7 fails at 2 AM.
• You will further develop our authorization model. Our clients range from individual owners to corporations with access at the organizational, portfolio, and building level. Building this correctly is real design work, not middleware configuration.
• You will integrate LLMs into backend workflows - primarily for document and data extraction - and make them reliable and testable.
• You will debug production. You will use Grafana dashboards and traces to figure out why something is slow or failing – whether in the app, the CI/build pipeline, or the infrastructure.
• You will identify overgrown or unnecessarily complex parts of the codebase and simplify them without being asked. Reducing complexity is part of the job, not a side project.

How We Measure Success

• You have delivered features end-to-end across schema, API, and background processing that our customers actively use.
• Our asynchronous processing (imports, extraction, analysis jobs) is measurably more robust: fewer silent errors, clear retry and recovery paths, and observability that tells us what's broken and why.
• You have taken ownership of at least one core area – e.g., the authorization model, job orchestration, or the LLM extraction pipeline – and are the go-to person for it in the team.
• You have elevated testing and developer experience in the backend to a new level, and other engineers are delivering faster as a result.
• You are a reliable sparring partner for Product and the team when we decide what to build and how.

What You Bring

• You are strong in Postgres. Schema design comes easily to you, you can read a query plan, and you know when the ORM is lying to you. Geospatial experience is a bonus, not a requirement.
• You are fluent in typed backend TypeScript. We work with NestJS and GraphQL daily; if your experience comes from a comparable stack, that's perfectly fine – you'll get up to speed.
• You have built real queue-based processing. Not "I once put a job in a queue," but designed workers, handled retries and partial failures, orchestrated multi-stage workflows.
• You have designed authorization for a multi-tenant product and know where the tricky parts are.
• You prefer an integration test against a real database over ten unit tests that mock everything. You write the test that catches the broken migration.
• You treat production as a debuggable system. You use dashboards and traces to figure out why something is slow, and infrastructure isn't someone else's problem for you.
• You communicate clearly – with engineers, product, and, if necessary, directly with customers.
• You speak German at a C1 level or higher and fluent English. Our clients are German banks and asset managers, and you will speak with them.

Bonus Points:

• AWS (S3, SES, ECS, RDS, …) and Infrastructure-as-Code.
• LLMs in production, especially document or data extraction – and the ability to make them reliable rather than just impressive.
• Data pipeline experience: staged, reproducible processing, dataset versioning.
• Mentoring or leading engineers – or the desire to grow into it.
• Debugging tooling and build problems, not just application code. Our toolchain is evolving rapidly and is occasionally non-standard.

Why Predium?

• Stage: ~40 people, well-funded, on the path to define the category. Early enough that much still needs to be built. Yes, it's sometimes messy – but that's the point. If you're looking for a finished playbook job, this isn't the right fit.
• Impact: You are not cog #100 in a company. What you build here is what our customers rely on every day. Your contribution is directly visible, in revenue and in the product.
• Ownership: We don't manage your time, we measure your results. You will be given responsibility and freedom – and the expectation to use both.
• Compensation: Competitive Salary, Bonus, and Equity participation. We want you to share in the success you help build.
• Team: Office-first in Werksviertel Munich, 4 days/week.

30 vacation days (incl. company holidays) · Wellpass · Deutschland-Ticket · Jobrad · Team Events

Find Jobs in Germany on Arbeitnow