I collaborated with Lettria on their GraphRAG solution, a generative AI platform designed for regulated industries (CSRD, ESG, compliance). My role spanned product improvement and feature design, working closely with their product and engineering teams.

UI Kit Foundation: Established the design system foundations (colors, typography, spacing, border-radius, and shadows) ensuring consistency from day one. As features evolved, I created components on-demand in close collaboration with developers, ensuring the system remained lean, adaptable, and technically feasible.

Chat History: Designed a multi-conversation system allowing users to maintain parallel discussions across different assistants and document sets.

Document-Based Q&A: Conceptualized the "ask on documents" feature enabling users to query specific document collections and receive LLM-generated answers grounded in source material.

Parsing Visualization: Created a transparency layer showing which document chunks were used to generate responses, building trust and auditability.

Permission System: Designed granular user roles (workspace, assistant, folder-level) for secure collaboration in regulated environments.

File Explorer Redesign: Overhauled the document/folder interface to improve scalability and readability as document volumes grow.

Throughout the project, I worked hand-in-hand with the front-end developer, reviewing edge cases together, anticipating implementation challenges, and ensuring designs remained feasible and consistent as features evolved.

Together, we shipped an intuitive, compliant-ready UI that balances advanced AI capabilities with enterprise UX expectations, adapted to the realities of a growing startup.