Wizzo
Back to projectsProject Overview
Designed and built an AI mentor product system that turns chat, connected work context, goals, and follow-up into actionable quests.
Revisit operating modelCurrent AI mentor interface
Current Wizzo product surface from the public beta, centered on chat, goals, voice, and quest follow-through.
Situation
•Intent-to-Action Gap
AI conversations often stop at advice. Wizzo reframes that moment as an execution problem: how to turn goals, blockers, deadlines, and recent progress into next steps a user can actually finish.
•Product Context
The public beta positions Wizzo as an AI adventure mentor that combines chat, voice, connected work context, achievement loops, and quest-based follow-through.
•System Context
The product needed to connect AI guidance with real work surfaces while keeping privacy, account controls, and product trust visible from the start.
Task
•Design an AI Mentor Loop
Create a product loop where chat can capture intent, remember execution context, suggest grounded next steps, and translate progress into quests.
•Connect Work Context
Support Drive, Gmail, Calendar, and technical workflows so the AI mentor can reason from the artifacts and deadlines that shape real follow-through.
•Make AI Actions Grounded
Pair conversation with tools for search, research, code execution, image generation, files, and voice so the assistant can move beyond generic coaching.
•Preserve User Control
Build account export, confirmed deletion, and privacy-aware controls into the product system so personal progress data remains manageable.
Action
1.Product System Architecture
Designed the core experience around main chat, execution-aware memory, quest tracking, achievements, and connector-backed work context.
2.Full-Stack Implementation
Built the product with Next.js, TypeScript, Postgres, Drizzle ORM, Neon, Vercel, OpenAI integrations, and Google workspace connectors.
3.AI Workflow Surfaces
Created mentor interactions for goal planning, progress reflection, grounded research, connected notes, and voice-supported follow-up.
4.Trust and Account Controls
Added export and confirmed deletion paths covering product, AI, notification, integration, ML-derived, and community records.
Current AI mentor interface
Current Wizzo product surface from the public beta, centered on chat, goals, voice, and quest follow-through.
Result
•Public Beta
Shipped a public beta with a marketing site, live web app, AI mentor interface, quest system, and connected work context.
•AI Product Proof
Demonstrated an AI-native product system that combines chat, memory, work connectors, tools, privacy controls, and habit-forming progress loops.
•Portfolio Relevance
Serves as proof of product design, engineering execution, AI workflow design, and operational thinking across a real shipped SaaS surface.
Beta product cockpit
Earlier beta interface showing the AI adventure mentor, quest status, and progress-oriented navigation.