• Own the roadmap for your product area. Define the vision, set priorities, and ship measurable outcomes. • Run discovery directly with customers and internal teams to validate real pain before committing engineering time. • Use data to frame problems, size opportunities, and judge whether what you shipped actually worked. • Partner with engineering and design from problem definition through launch and iteration. • Translate strategy into clear, sequenced delivery, and keep stakeholders aligned on what's coming and why. • Balance speed and trust. Ship fast where you can, but hold the line on quality, privacy, and compliance.

• Partner directly with enterprise customers to deeply understand their workflows, challenges, and goals. • Translate complex customer needs into clear, structured AI solution plans using Unframe’s platform. • Own end-to-end delivery of AI-powered projects alongside engineers and AI specialists. • Write detailed product specs and define prompt/context structures that drive solution behavior. • Balance speed, scalability, stability, and customer impact when making product decisions. • Identify repeatable patterns and collaborate with the Platform team to improve and scale reusable capabilities. • Act as the voice of the user throughout the entire product lifecycle.

• Partner directly with enterprise customers to deeply understand their workflows, challenges, and goals. • Translate complex customer needs into clear, structured AI solution plans using Unframe’s platform. • Own end-to-end delivery of AI-powered projects alongside engineers and AI specialists. • Write detailed product specs and define prompt/context structures that drive solution behavior. • Balance speed, scalability, stability, and customer impact when making product decisions. • Identify repeatable patterns and collaborate with the Platform team to improve and scale reusable capabilities. • Act as the voice of the user throughout the entire product lifecycle.

• Own an AI agent end-to-end: Drive the full lifecycle — discovery, prioritization, spec, delivery, iteration — for 1–2 agents at a time, from first hypothesis to live product. • Set and hold the delivery cadence: Define what "done" looks like, track progress, and unblock the team — without waiting for someone else to push. • Talk to users, constantly: Regular user calls are non-negotiable. • Partner with domain experts: Work with cross-functional teams (e.g., Financing for banking agents) to acquire the context and co-define features. • Help scale agent practices: Contribute to the patterns, handover models, and playbooks that will eventually let AI agents live across Qonto's product organization.