Full-Stack Engineer - Data, Infrastructure & AI

About Agent F

Agent F is building an autonomous finance function: software that takes on the work of keeping a company’s finances in order. That means bringing information together from different systems, recording transactions correctly, preparing useful reports, and helping people understand what needs their attention.

Today, much of this work depends on spreadsheets, repeated data entry and people checking one system against another. Our ambition is to make that work increasingly automatic, while keeping every number explainable and every consequential action under clear control.

We are building towards a complete business management system for growing companies. These founding roles will shape both the product and the way we build it.

About the role

You will be a full-stack engineer, shipping customer features across the interface, application services and database. Your particular strength will be the data and infrastructure that make those features dependable: how records are stored, how imports and background jobs run, and how information moves between Agent F and a customer’s other tools. You will help build and operate the whole product, with deeper ownership of this foundation.

This is hands-on full-stack engineering. A typical piece of work could involve extending a data model, implementing an application service, connecting it to a screen, testing it against a real database and following the release into production. You will also investigate slow reports, failed imports and jobs that stop halfway through, then improve the system so it handles those situations reliably.

That foundation also supports our AI work. As we add document extraction, connected customer knowledge and software that takes actions for customers, you will build the services and data access those features need. The immediate priority is dependable software that customers can use, with more capable automation developed around specific tasks.

What you will build and improve

• Application services and customer features. Build the logic and application programming interfaces (APIs) that connect records, business rules and screens. Take features through implementation, testing, release and production support.
• The database foundation. Design clear data models, enforce relationships and transaction boundaries, and improve slow queries. Manage schema changes through reviewed migrations, preserve financial history and keep each customer’s data isolated.
• Imports and integrations. Connect accounting tools, business systems and spreadsheets to a consistent model. Validate incoming records, handle changing source formats, and check that imported totals and reports agree with the source.
• Reliable background processing. Build scheduled jobs and workers that process work outside a user request. Handle retries without creating duplicates, resume interrupted runs, and make progress and failures visible to the team.
• Performance and recovery. Diagnose expensive queries, processing bottlenecks and partial failures. Test changes against real databases, plan for growing workloads, and build procedures to recover processing safely with the platform engineer.
• Practical AI automation. Add features such as extracting invoice fields or preparing an order from approved business data. Give models access through controlled services, check the result and require approval where needed. Develop connected customer knowledge when a feature needs those relationships, and measure whether it improves the result.

For example, a large customer import stops halfway through and the same records arrive again on the next run. You would make processing resume safely, prevent duplicate records, verify the resulting totals and expose the job’s status in the product. You would also investigate whether database queries or worker capacity caused the failure and ship the necessary fix.

What you bring

You have seven or more years of software engineering experience, including substantial ownership of production applications and data systems. You have designed relational data models, built services and processing pipelines, and stayed responsible for them after release. You are comfortable delivering across the stack and using AI tools in your daily work.

You can explain how you diagnosed a slow query, recovered a failed job or shipped a database change safely. You understand what happens when two requests arrive at once, a dependency becomes unavailable or a process restarts halfway through its work. You can follow a problem from a source record through the application to the result a customer sees.

Core experience

• Full-stack development with TypeScript, Node.js, React and PostgreSQL, or closely comparable technologies.
• Relational data modelling, SQL, query optimisation, indexes, transactions and reviewed database migrations.
• Application services and APIs, including permissions, validation and handling concurrent requests.
• Data imports, integrations and background jobs, including queues, safe retries and recovery from partial failures.
• Production debugging, service monitoring and performance work on systems used by real customers.
• Automated tests against real databases and other service dependencies.
• Practical use of AI in software, with an understanding of how to check outputs and control access to customer data.
• Daily use of AI coding agents, with direct responsibility for code review and verification.
• A computer science or computer engineering degree, or equivalent practical depth.

Useful additional experience

Financial or operational systems where correctness and traceability matter; larger data migrations; cloud storage and container-based workloads. Experience putting document extraction or features using large language models (LLMs) into production would be valuable. Knowledge graphs, which connect related company records, and privately hosted or adapted models are useful as we develop that part of the product.

What early progress could look like

In your first months, you would take ownership of an existing service or data flow, understand how it runs in production and ship a customer-facing improvement through it. You would address a concrete reliability or performance problem, such as a slow report, a fragile integration or a job that cannot recover cleanly after interruption.

We would look for working features, faster and more predictable processing, fewer failed or manually repaired jobs, and useful tests and monitoring. An early automation feature could follow where the data and services are ready and its benefit can be measured.

Operational excellence and customer commitments

You take pride in a dependable application backbone: processing completes, records reconcile and customers can trust the results. When a job fails or figures disagree, you take ownership of the investigation and recovery. Missing transactions, duplicate entries or unexplained balance differences can be financial emergencies even while every service appears available. You assess the impact, contain affected processing, trace the cause and verify the repaired result before treating the incident as resolved.

Customer commitments shape how we work. You are willing to interrupt planned work, respond promptly and stay engaged when an urgent issue puts a service-level agreement (SLA), onboarding commitment or financial deadline at risk. That requires flexibility beyond normal working hours as part of agreed on-call coverage and coordinated incident response. You keep people informed and see the issue through to a verified recovery or an explicit handover. Afterwards, you improve the system so the same emergency becomes less likely.

How you will work

This is a senior founding role in a small team. You will own the application and data infrastructure for your work, and carry a full share of customer feature delivery. You will collaborate across infrastructure, operations and product design to deliver complete customer outcomes. You will use coding agents to implement changes and remain responsible for reviewing, testing and operating the result.

Everyone shares responsibility for the whole product and the customer outcome. You will pick up the most important work or whatever is on fire, including tasks outside your specialty. You take initiative, bring in the right expertise and stay involved until the problem is resolved. Customer impact and urgency determine where you focus.

The role is full-time and in-person in Berlin. German is not required.

Find Jobs in Germany on Arbeitnow

We show an excerpt and a link, and the full text when the source allows it. Brands and logos shown belong to their owners.

Source: Statistisches Bundesamt (Destatis), table; dl-de/by-2-0; StarIn's calculation

Listed on a job portal Nov 07, 2026

"visa support" · "relocation support"