Easy-to-manage touch screens
Free, open-source touch screens. Guided USB setup, a drag-and-drop editor and local control. Made for Home Assistant.
翻訳待ち
No updates detected yet. GitHub releases will appear here.
Loading the full README. The saved excerpt is shown below.
Stop Guessing. Start Winning Tatkal.
An AI travel agent that prepares you before the Tatkal window opens. It never
books, logs in, or automates anything — it makes you dramatically better
prepared, then hands you off to the official IRCTC site.
Built for the Build What Moves India hackathon (OpenAI + Varun Mayya).
Everything is designed for Manoj Sharma, 54, Mumbai — books Tatkal 3–5×/year,
not tech-savvy, usually relies on his son or a travel agent. His goal:
"I need to reach Delhi before 8AM tomorrow."
A single continuous story across 8 screens:Read the full README ↗Stop Guessing. Start Winning Tatkal.
An AI travel agent that prepares you before the Tatkal window opens. It never books, logs in, or automates anything — it makes you dramatically better prepared, then hands you off to the official IRCTC site.
Built for the Build What Moves India hackathon (OpenAI + Varun Mayya).
Everything is designed for Manoj Sharma, 54, Mumbai — books Tatkal 3–5×/year, not tech-savvy, usually relies on his son or a travel agent. His goal: "I need to reach Delhi before 8AM tomorrow."
A single continuous story across 8 screens:
The AI does the two things it is genuinely good at:
Every number and the strategy itself stay grounded in mock data via a
deterministic local planner (lib/planner.ts), so GPT can never invent a fake
confirmation figure.
Dual-path resilience: with OPENAI_API_KEY set, the /api/plan route uses
GPT (gpt-4o-mini). Without a key — or if the call fails — it falls back to the
local planner and produces the same shape of expert output. The demo never
breaks in front of judges.
npm install
npm run dev
Open http://localhost:3000.
To enable the live GPT path:
cp .env.example .env.local # add your OPENAI_API_KEY
Next.js 16 · React 19 · TypeScript · Tailwind v4 · Framer Motion · Lucide. Local JSON data, no database, no auth, no backend beyond one API route.
app/ layout, page orchestrator, /api/plan (GPT + fallback)
components/ ui primitives (button, card, probability ring), brand chrome
features/ one folder per screen (hero, goal, thinking, strategy, …)
lib/ planner (intelligence), ai (dual-path), i18n (EN/HI), journey state
data/ trains, stations, split routes, passengers, mission script (JSON)
hooks/ countdown, speech
types/ domain model