Limetax is building the operating system for the €20B German tax advisory industry. We acquire established tax firms and transform them into high-performing, technology-enabled businesses. We combine proprietary software, streamlined operations, and a strong group infrastructure to unlock value that traditional firm owners can't access alone.
Our founding team brings together backgrounds in investment banking, venture building, and applied AI. We're a small, fast-moving team backed by top-tier investors. Our mandate is clear: acquire firms, make them measurably better through technology, and scale. Every person we hire has real ownership and direct impact on outcomes.
The RoleDesign, build, and operate applied AI systems that meaningfully reduce manual work in tax advisory.
Limetax sits on large volumes of real, high-quality production data: transactions, documents, client histories, and domain knowledge across multiple firms. The AI systems you build will define which parts of tax work can be automated safely and at scale, how humans and machines collaborate in high-stakes workflows, and how trust, explainability, and reliability are embedded into AI from day one.
What You'll OwnBuild and operate document intelligence systems on real tax documents.
Design and implement RAG systems over tax law, internal knowledge bases, and client histories.
Build and maintain data pipelines that power ML and AI workflows.
Develop prediction and suggestion systems, e.g. booking proposals and anomaly detection.
Own evaluation, monitoring, and continuous improvement of AI systems in production.
Work closely with engineers and domain experts to translate real workflows into AI-driven solutions.
Have shipped ML/LLM-based systems into real production environments.
Think like an engineer first. Reliability, observability, and failure modes matter.
Are curious about applying AI safely and effectively in regulated domains.
Enjoy working hands-on with messy, real-world data.
Are comfortable in a TypeScript-first stack (Python welcome).
Access to real production data from day one. Transactions, documents, client histories across multiple firms. No synthetic demos.
AI systems with immediate operational impact. You'll see your work change how tax work gets done within weeks.
Tight coupling between data, product, and real workflows. You operate as an engineer in a domain that genuinely needs what you're building.
€4M+ seed funding and founders with a proven track record. You join at the moment things are being built for real.
Flat hierarchy, high trust, direct access to founders. No bullshit, no layers between your work and its impact.
Competitive compensation and early equity.
We read every application and move quickly when we find the right person. There will be a few conversations and a practical case. We only make offers when we're confident it's the right fit for both sides.
Skills Required
- Shipped ML/LLM-based systems into production
- Strong engineering focus on reliability, observability, and failure modes
- Experience building and operating document intelligence systems
- Experience designing and implementing RAG systems over knowledge bases
- Build and maintain data pipelines that power ML and AI workflows
- Develop prediction and anomaly detection systems
- Comfortable in a TypeScript-first stack
- Experience with Python
- Ability to work hands-on with messy, real-world data and collaborate with domain experts
- Interest in applying AI safely in regulated domains with emphasis on trust and explainability
What We Do
Limetax is an AI-enabled roll-up in the accounting space, focused on building the operating system for the €20B German tax advisory industry. The company acquires established and profitable tax firms, transforming them into high-performing entities to modernize how CFO and tax services are delivered to German small and medium-sized enterprises (SMEs).
