Vordo AI
Designed an AI note-taking system that thinks in tasks, not just text.

Every note-taking and task management app has the same problem: you still have to do the organising. You type a task, then manually decide which list it goes in, which priority it gets, which date it belongs to. The cognitive overhead of maintaining the system is almost as high as the work itself. Voice input exists but it just transcribes - it doesn't think.
The core insight was that tasks naturally fall into different time horizons - and AI could classify them automatically from how you describe them. A project that takes weeks is fundamentally different from a task you'll do today, which is different again from something you want to remember years from now. I designed a three-tier system: macro tasks (multi-day projects), daily tasks (today's to-do), and memory (long-term records like when you changed your car battery - searchable forever). The key product decision was making voice the primary input, not a secondary feature, and using ElevenLabs for the voice layer with AI classification happening in the background before anything hits the UI.
Full product strategy, UX design, and PRD for Vordo AI. Designed the three-tier task classification system, voice input flow, custom calendar integration, and search interface. The product reached a complete design and specification stage but was not developed - the concept was validated at the design level.
Delivered a complete product design and go-to-market strategy. The three-tier classification model and voice-first architecture were validated through design review. The product did not reach production - this is a product strategy and design case study.
Product designed in Figma. AI classification architecture designed around ElevenLabs for voice input and an LLM layer for intent classification and task routing. Custom calendar designed to surface time-horizon context alongside standard scheduling. Full PRD written covering user flows, edge cases, and the classification logic.