Skillpath

• 

  Dashboard

• 

  Explore paths

• 

  generated path

• 

  Landing Page

What problem does SkillPath solve and why I built it?

Every time someone decides to learn something new — whether it's UI/UX design, Python, or public speaking — they face the same invisible wall: where do I actually start? YouTube has thousands of tutorials. Google returns hundreds of articles. Reddit threads contradict each other. The result is decision paralysis, and most people quit before they begin. I built SkillPath because I experienced this myself. I wanted to learn web development, spent three weeks jumping between random resources, made no real progress, and gave up. What I needed wasn't more content — I needed a structured, personalized plan that told me exactly what to do this week, not someday. SkillPath solves this by turning a single sentence — "I want to learn UI/UX design in 3 months as a beginner with 30 minutes a day" — into a complete, actionable, week-by-week learning roadmap in seconds.

How I structured conversations with MeDo

MeDo was at the center of every layer of this app, not just one feature. The first conversation I had with MeDo was describing the entire app vision — pages, features, database schema, design system, and AI behavior — in one detailed prompt. MeDo scaffolded the full-stack app from that single conversation: frontend pages, backend routes, database tables, and authentication all came together in the first session. From there, I used follow-up conversations to refine specific features. I asked MeDo to improve the onboarding flow with sliding step animations, then asked it to make the roadmap timeline more visual with week status indicators. Each iteration was a natural back-and-forth — I described what felt off, MeDo fixed it and explained what it changed. The most powerful conversation was building the AI chat panel inside the roadmap view. I told MeDo: "I want users to be able to talk to the app about their current week — ask for simpler explanations, request a quiz, or say they only have 10 minutes today — and get a helpful, context-aware response." MeDo built the full multi-turn chat system with roadmap context passed into every message, so the AI always knows which week the user is on and what resources they've already seen.

The most impressive feature MeDo helped create

The personalized roadmap generation engine is what I'm most proud of. When a user completes the onboarding flow — specifying their goal, experience level, daily time available, and learning style — MeDo helped me build a backend function that sends all of that context to an AI model and receives a fully structured JSON roadmap in return. Each week has a theme, 3–5 curated resources with titles and types (video, article, or project), and an estimated time commitment. What makes this technically impressive is that the roadmap isn't generic. A beginner who has 20 minutes a day gets a slower, more foundational plan than an intermediate learner with 2 hours a day pursuing the same goal. The AI adapts the pacing, depth, and resource types based on the user's learning style preference. MeDo helped me write the prompt engineering logic and the JSON parsing layer that feeds directly into the database and renders on screen — all within one conversation.

How I used plugins or APIs to extend functionality

SkillPath integrates several external capabilities to go beyond a basic AI wrapper: Google OAuth was connected for one-click sign-in, removing friction from the moment a user lands on the app. MeDo helped me wire the authentication flow into the user database in minutes. The Explore page pulls public roadmaps from the database and lets any user — even without an account — browse what others are learning. This turns SkillPath into a community tool, not just a personal one. MeDo built the public/private roadmap toggle and the sharing link system that generates a unique URL for every roadmap. PDF export lets users download their roadmap as a formatted document to reference offline. MeDo integrated a PDF generation library and styled the output to match the app's visual design. Finally, the suggested prompt chips in the AI chat panel — "Simplify this week," "I only have 20 minutes today," "Give me a quiz" — were built using MeDo's suggestion to pre-populate common user intents, dramatically improving the experience for first-time users who don't know what to ask

Built With

• medo