Inspiration
Traject3 started from a simple observation: most people know they want a better future, but they have no idea what direction to run, what skills matter, or where to even begin. Tools today either overwhelm users with options or gatekeep knowledge behind jargon, prerequisites, and long learning curves.
I wanted to build something that gives people a moment of clarity—a spark—where one sentence about who they are can be converted into a clear, believable, achievable future identity. And from there, help them take their first meaningful step, not just explain concepts or dump resources. Traject3 was inspired by the idea that: Your trajectory should be personal, contextual, and instantly actionable. People don’t need more courses—they need a path.
What Traject3 Does
Traject3 is a future-skills activator. Users type a simple statement (“I want to change careers”, “I love design”, “I’m curious about AI”, etc.), and the system generates: A future identity they can grow into A skill roadmap tailored to their background Three missions they can start today Real-world tools and resources matched to their level A follow-up personalization flow that adapts the path even further In one click, Traject3 turns ambiguity into direction, and direction into action.
How I Built It
I designed Traject3 to be both fast and flexible:
Core Components
Cursor + AI-driven code workflow to prototype rapidly Next.js frontend for speed, routing, and clean UI Custom LLM prompts to synthesize identities, paths, and missions A modular “Trajectory Engine” layer that converts raw user text into structured outputs Lightweight database for saving profiles and sessions (SQLite/Airtable for MVP)
Architectural Approach
A user input hits the Trajectory Engine. Their background → embeddings → match patterns across known career transitions. The system generates a future identity + roadmap. Missions and steps are dynamically chunked by difficulty and starting point. UI renders everything in a clean, actionable format. Everything was designed to be lean, fast, and demo-ready.
Challenges I Faced
Designing a trajectory system that doesn’t give generic answers took several iterations of prompt engineering and output-format constraints. Avoiding information overload—the goal was clarity, not dumping links or resources. Creating missions that feel truly “startable today” rather than abstract advice. Building a system that works even with minimal user input (“I like tech”). Balancing personalization, simplicity, and actionability was the hardest part.
What I Learned
The difference between a good AI tool and a great one is friction—remove friction and the product becomes addictive. Users don’t need more knowledge; they need direction. Thinking like a career counselor, not an engineer, produced better results. Minimal input → maximum clarity is the new UX frontier, especially in AI. Small, well-designed missions create more user momentum than long roadmaps.
What’s Next
Traject3 will expand into: Personalization engine with multi-step intake Modular learning arcs that evolve as users complete missions A dynamic “Trajectory Map” visualization Project-based progression paths (instead of just reading or watching) Integrations with actual resources, mentors, communities, and tools Eventually, Traject3 aims to become a human potential GPS—guiding users from where they are to where they’re capable of going.
Log in or sign up for Devpost to join the conversation.