Founding Engineer

About The Role

This is a founding engineering position at an early-stage startup focused on creating AI systems that minimize manual tasks within tax advisory and accounting firms. The role involves working closely with actual production workflows and domain experts to evolve AI solutions from experimental prototypes into dependable, auditable systems that manage critical financial data at scale.

Key Responsibilities

• Develop and manage document intelligence systems for extracting, classifying, and processing complex accounting and tax-related documents.
• Design and implement retrieval augmented generation (RAG) systems integrating tax laws, internal knowledge bases, client histories, and firm-specific data.
• Create AI-powered agent workflows automating multi-step accounting and tax processes.
• Build prediction and suggestion tools addressing booking proposals, anomaly detection, classification, and intelligent recommendations.
• Maintain data pipelines and infrastructure to support robust AI and machine learning workflows.
• Establish frameworks for evaluation, monitoring, observability, and feedback to measure system performance and identify failure points in production.
• Iteratively enhance AI capabilities using real-world performance insights.
• Collaborate with engineering teams, operations personnel, and tax experts to incorporate domain knowledge into AI solutions.

Candidate Profile

• Minimum of 3 years hands-on experience deploying machine learning, large language models (LLMs), or AI systems into production environments.
• Proficient developing and operating LLMs, retrieval systems (RAG), or AI agents within live applications.
• Expertise in building document intelligence workflows including extraction, classification, and comprehension of unstructured documents.
• Strong software engineering skills emphasizing testing, reliability, observability, and maintainability of AI systems.
• Experience designing evaluation methodologies and monitoring tools for productive AI models.
• Proven ability to transition AI capabilities from prototypes to iterative, reliable production releases.
• Demonstrated collaborative experience partnering with domain experts to translate complex processes into AI-driven systems.
• Background in domains demanding high reliability and auditability, such as finance or tax, is advantageous.
• Additionally valued experience includes working with agentic workflows, anomaly detection, classification, and recommendation engines in production.

Work Location

This position requires physical presence on-site in Berlin, Germany.