Senior Product Manager (m/w/d)

About This RoleAI processing…

Key Responsibilities

• 1
  Ship AI Features That Move Numbers: You own AI features end-to-end – document recognition, classification, auto-drafting of tax responses, RAG over client files. You build the eval frameworks, run the prompt iterations, and make the build-vs-buy calls on models and tooling. No hand-offs, no excuses.
• 2
  Get Obsessed With the User: You run discovery the hard way: interviews, shadowing sessions, raw feedback synthesis. You go deep enough to tell real problems from assumed ones – and become the sharpest voice for the user inside your product line.
• 3
  Own the Roadmap, Defend Trade-Offs: You own your product line's roadmap end-to-end. Prioritization across business value, technical feasibility, and regulatory fit. You communicate trade-offs sharply to Engineering, Tax Advisory, and Founders – and you defend your calls with data.
• 4
  Hold the Bar on Design: You set a high visual bar and protect it. Figma fluency at the review-and-sketch level: you draft flows, critique components, push back on anything that doesn't meet the standard. The best tax UX is invisible – make ours invisible.
• 5
  Run on Data, Not Vibes: You define the KPIs that matter – Time-to-Filing, automation rate, client satisfaction, advisor efficiency – and you steer every iteration from dashboards. Gut feel doesn't ship features here.
• 6
  Push the AI Frontier: You don't follow the AI conversation – you drive it. Evals, prompt iteration, cost/quality trade-offs in production. You experiment relentlessly with what's possible and bring the edge into the product.
• 7
  Be the Connective Tissue: You sit between Engineering, Tax Advisory, and Clients and make the org work faster because of it. First-principles thinking

Requirements

• Product Experience: 3–6 years in product management in B2B SaaS, FinTech, Finance, Tax, or a comparably complex environment. You can point to features you owned from vision to measurable impact – not features you helped with.
• AI Track Record: At least one AI/LLM feature shipped in production – with real evals, prompt iteration, and ownership of cost/quality trade-offs. Solid grasp of modern LLM stacks: RAG, prompt engineering, fine-tuning vs. in-context learning, vector stores. ChatGPT prompts don't count – we want AI-native workflows you actually built.
• Client Empathy That Translates: Genuine empathy for both sides – the advisor under deadline pressure and the client who just wants certainty – and the discipline to turn that empathy into product decisions, not backlog items.
• Eye for Design: Figma fluency at the sketch-and-review level. Strong opinions on what good UI looks like and zero tolerance for sloppy interfaces. Excitement for Claude Design is a plus.
• AI Enthusiasm: You use AI products daily, you critique them publicly or privately, and you have sharp views on where they're going. Hands-on familiarity with Claude Cowork/Design/Code, GPT, Gemini or similar is the baseline.
• Self-Implementation: You don't wait to be unblocked. You prototype, test prompts, wire up integrations, hack together what's needed. If it's blocked, you unblock it.
• Structured Thinking: MECE, first principles, hypothesis-driven. You compress ambiguity into clear decisions – not into more slide decks.
• Communication: Excellent English (C2). German is a strong plus.
• Domain Background: Experience in Finance, Tax, Accounting, or adjacent domains is a strong plus – as is familiarity with accounting/finance software stacks, DATEV, ELSTER, or regulated data workflows.

Perks & Benefits