Product Manager with AI

Wymagania: - 3-6 years of product management experience in SaaS, analytics, startups, or AI products - Vibe-coding enthusiast: You can sit down and code up a working prototype (HTML/CSS/JS) to demonstrate UX flows—you don't need perfect production code, but you can make things work - Fluent with GenAI: Deep understanding of LLM capabilities, limitations, and practical applications; you know the difference between GPT, Claude Sonnet, and smaller models, and when to use each - AI workflow design: Experience designing products that incorporate AI components—prompt engineering, context windows, chaining, error handling - UX/UI fundamentals: Solid grasp of good design practices, information architecture, and user-centered design—enough to create wireframes and evaluate interface decisions - Rapid prototyping mindset: Comfortable working without pixel-perfect designs; can translate concepts into working mockups independently - Clear, concise communication: Can explain complex technical concepts simply, write crisp requirements, and facilitate productive conversations - Highly self-organized: Thrives in ambiguous, fast-moving environments; doesn't need handholding or detailed processes to stay productive - Comfortable with Azure DevOps (or similar tools) for user story management and backlog tracking - Available during EU/US business hours for team collaboration (remote-first culture) Nice to have: - AI-assisted coding tools: Experience with Claude Code, Cursor, Roo Cline, Windsurf, or similar AI coding assistants—you can ship faster by leveraging these tools - MCP (Model Context Protocol) knowledge: Understanding of MCP servers and how to leverage them in AI application architectures or own workflows - Audit domain knowledge: Background in Big 4 accounting, financial auditing, or regulatory compliance—helps you understand user context faster - Technical depth: Ability to understand API design, data models, and technical architecture discussions (not required to write production code, but helpful for collaboration) - Previous experience in B2B enterprise products with complex stakeholder landscapes - Familiarity with modern frontend frameworks (React, Vue, etc.) beyond basic HTML/CSS/JS Codzienne zadania: - Distill the signal from noise: Parse verbose, domain-specific user feedback to identify the real underlying problems - Triage incoming streams: Manage user feedback, bug reports, and logic change requests across 3 applications simultaneously - Write crisp user stories: Create clear, actionable Azure DevOps user stories with acceptance criteria that developers can immediately execute on - Prioritize ruthlessly: Make fast prioritization decisions in daily prioritization sessions—no time for lengthy backlog grooming ceremonies - Propose solutions: Don't just surface problems—bring solution hypotheses backed by user understanding and technical feasibility - Prototype rapidly: Sit down and vibe-code frontend logic/workflows to demonstrate how features should work (we don't have dedicated designers for every iteration) - Create wireframes/mockups: Translate ideas into visual artifacts that communicate UX/UI intent clearly - Iterate based on feedback: Test assumptions quickly, adjust, and move forward without getting stuck in analysis paralysis - Be the dev team's partner: Answer questions swiftly, unblock developers, clarify edge cases on the fly - Facilitate smooth delivery: Keep things moving—your success metric is reliable, on-time delivery - Engage stakeholders when needed: Calibrate with auditors and business stakeholders to validate domain logic and priorities - Daily standups & sporadic syncs: Participate in lightweight, high-signal team rituals—no ceremony for ceremony's sake - Stay current with AI capabilities: Understand the latest LLM models (GPT-5, Claude 4.5, Gemini 2.5, etc.) and their tradeoffs - Design AI workflows: Think through prompt engineering, context management, error handling, and user feedback loops - Optimize for reliability: Balance innovative AI features with system uptime and predictable behavior