ArchiTech

Inspiration

Creators today spend more time thinking of what to post than actually creating.
We were inspired by the rise of AI copilots and the new UI patterns they bring.
Our goal was to help TikTok creators save time, stay consistent, and keep up with trends—while showcasing how Lynx can power modern, cross-platform UIs in the AI era.

What it does

Generates AI-powered captions and hashtags based on uploaded videos.
Provides personalized suggestions by learning from a creator’s past posts.
Shows inline analytics like trending hashtags and engagement confidence.
Offers an AI chat panel where creators can refine captions in real-time.
Built with a TikTok-style UI (swipe cards, confirmation flows, chat interface).

How we built it

Frontend: Lynx + ReactLynx with Tailwind for styling and TikTok-style interactions.
Backend (Python): FastAPI + SQLite + SQLAlchemy to manage creators, preferences, and analytics.
Backend (Node.js): Express + Multer for video uploads and mock AI caption generation.
AI Layer: Planned integration with OpenAI / Vercel AI SDK (currently mocked).
Database: SQLite seeded with creator data, preferences, and memories for personalization.
Tools: Swagger for API docs, GitHub for version control, VS Code & Docker for development.

Challenges we ran into

Learning how to build with Lynx, since it’s a new framework and quite different from typical React or web dev flows.
Adapting to mobile development with Lynx Explorer, which was new for most of us.
Our solution currently lacks deep integrations (e.g., real AI APIs, TikTok APIs) and needs more polish.
Designing a smooth UI/UX within Lynx was tricky, and we identified areas for major improvement.
Managing cross-platform integration (making it work seamlessly on web and mobile) remains an ongoing challenge.

What's next for ArchiTech

Improve the UI/UX for a more polished TikTok-like feel with smoother interactions.
Add stronger cross-platform support so the app works seamlessly on both web and mobile.
Replace mock APIs with real integrations (OpenAI/Vercel AI SDK for captions, TikTok APIs for trend detection).
Enhance personalization engine with smarter analytics and memory tracking.
Expand beyond TikTok into a multi-platform creator copilot.