You will join our AI & Automation team, with a dedicated product manager and a tech lead, build, and ship AI-powered features that fundamentally change how accounting works
The goal is exception-driven accounting
where AI handles the majority of transactions automatically, and humans focus on review, control, and complex cases
You will build production systems that process real financial data for real businesses, with high standards for correctness, auditability, and transparency
Agentic workflows and automation
Build intelligent automation that suggests, matches, and processes accounting transactions
Design confidence scoring, decision thresholds, and human-in-the-loop approval flows that determine when automation can proceed safely and when human review is needed
AI runtime services
Develop and maintain the AI services (Python / TypeScript) that power classification, matching, and suggestion logic
Evaluate and integrate LLMs for specific accounting use cases, including RAG pipelines and tool-calling architectures
Assistant experience
Contribute to BHB's AI assistant that helps users with accounting queries, task execution, and contextual guidance within the platform
Learning loops and metrics
Build feedback mechanisms so the system continuously improves
Track automation rates, suggestion acceptance, and error patterns. Turn user corrections into training signals
Data foundation for AI
Work with structured transaction and line-item data to ensure high-quality inputs for all AI features
Help build the semantic context engine that gives agents the right context for decisions
Requirements
Between 3 and 5 years of professional experience in software development with a strong focus on AI-powered product features
Hands-on experience building production AI systems
not just notebooks and prototypes, but services that run reliably, handle edge cases, and are observable in production
Proficiency (having at least 3 years of hands-on experience) with Python and/or Typescript and familiarity with modern AI/ML tooling (LLMs, RAG, embeddings, vector stores, prompt engineering)
Understanding of agentic architectures
tool-calling, orchestration patterns, confidence-based routing, and human-in-the-loop design
Strong software engineering fundamentals and experience
clean code, API design, testing, CI/CD, microservices and the discipline to ship safely in a domain where correctness matters
Comfort with ambiguity and ownership
this is a growing team. You'll shape how AI works at BHB, not just execute tickets. You should be proactive by figuring out the right approach, not just implementing a spec
Nice to have: Experience in fintech, accounting, or compliance-heavy domains
Familiarity with German accounting concepts (SKR03/SKR04, GoBD, DATEV)
Experience with MCP (Model Context Protocol) or similar tool-integration patterns