Founding Engineer
About the Role
This is a founding engineering role at an early-stage startup building AI systems for the tax advisory and accounting industry. You will work directly with real tax firms and their production workflows, turning complex domain knowledge into reliable, production-grade AI solutions that meaningfully reduce manual work for tax professionals.
What You'll Do
Build and operate document intelligence systems that extract, classify, and process real accounting and tax documents.
Design and implement retrieval and knowledge systems (RAG) across tax law, internal knowledge bases, client histories, and firm-specific data.
Build agentic workflows and AI-driven automation for complex, multi-step accounting and tax processes.
Develop prediction and suggestion systems for use cases such as booking proposals, anomaly detection, classification, and intelligent recommendations.
Build and maintain data pipelines and infrastructure to power reliable AI and ML workflows.
Design evaluations, monitoring, observability, and feedback loops to track system performance and failure modes in production.
Take AI capabilities from experimentation through deployment and iterate continuously based on real-world performance.
Collaborate closely with engineers, operators, and tax professionals to translate domain workflows into AI-driven solutions.
What We're Looking For
3 or more years of hands-on experience shipping ML, LLM, or AI-based systems into real production environments.
Hands-on experience with LLMs, retrieval systems (RAG), or AI agents in production.
Experience building document intelligence systems, including extraction, classification, and understanding of unstructured documents.
Strong software engineering discipline: testing, reliability, maintainability, and observability in AI systems.
Experience designing evaluation frameworks and monitoring systems for AI in production.
Experience designing and implementing data pipelines for ML and AI workflows.
Track record of taking AI capabilities from prototype through production with iteration cycles.
Experience in high-stakes domains where reliability and auditability are critical.
Strong product intuition and ability to turn vague problems into pragmatic technical solutions.
Prior exposure to accounting, tax, or financial domain workflows is a strong plus.
Experience building agentic or multi-step AI automation systems is a plus.
Location
On-site in Berlin, Germany.
