Crafting experience...
6/12/2026
Built At
Progress x GitNation
Hosted By
GitNation
Talks & workshops by core teams and top engineers.
Subtitle: Every room is a channel. Every session is a broadcast.
Demo Video: https://youtu.be/iy1dgo9XXC0
Live Demo: https://tejasvi-nagendrappa.github.io/conference-cast/
GitHub Repository: https://github.com/tejasvi-nagendrappa/conference-cast
You pay β¬500 to attend a conference. Five talks run simultaneously. You pick one at random, sit through it, and later find out the talk in Room B was exactly what you needed.
That moment of "I missed the right talk" is something every developer has felt:
For Attendees: Remote and in-person attendees face the same problem with five simultaneous streams and zero guidance.
For Organisers: Events are run completely blind. There is no real-time view of which rooms are packed, where Q&A is spiking, or where they are about to lose an audience.
Multi-track conferences are fundamentally an information problem.
ConferenceCast applies a TV broadcast metaphor to tech conferences. Attendees get an AI-powered programme guide, while organisers gain access to a real-time control room dashboard.
AI-Personalised Lineup: Sessions are scored 0β99 against your track interests and experience level, complete with a plain-English "why this matches you" explanation.
LIVE Tab: Displays all currently broadcasting sessions in a grid. Click any card to expand an inline video player and a live Q&A panel without leaving the page. The most-watched session automatically receives a π₯ Trending badge.
UPCOMING Tab: Sessions are grouped by time slot with live countdown timers. A toast notification fires 2 minutes before any bookmarked or relevant session starts.
GUIDE Tab: A full horizontal Electronic Programme Guide (EPG) timeline (09:00β13:00), colour-coded by channel, featuring a live NOW indicator. Ended sessions stay visible so in-person attendees can easily find missed tracks and watch recordings later.
Real-Time KPIs: Monitored via trend arrows tracking total live viewers, average room capacity, and Q&A activity rates.
"Now Broadcasting" Strip: Displays speaker avatars with live pulse dots, alongside per-channel viewer counts and room fill rates.
Live Viewership Chart: An area chart mapped across all 5 channels that updates every 3 seconds.
Alert Feed: Generates capacity warnings, Q&A spikes, and low-attendance flags so organisers can act before a room overflows or a talk loses its audience.
Event Pain Point | ConferenceCast Solution |
"I don't know which of 5 talks to watch." | AI match score + plain-English reasoning |
"I missed a session I would have loved." | EPG timeline shows ended sessions to watch later |
"I can't tell which room has the energy right now." | π₯ Trending badge placed on the most-watched live session |
"I forgot a talk was starting soon." | Toast notification sent 2 minutes before start time |
"Our biggest room is overflowing." | Control Room capacity alerts triggered in real time |
Framework: React 18 + TypeScript
Build Tool: Vite
State Management: Zustand
Routing: React Router v6 (HashRouter)
Hosting: GitHub Pages
Chart Core: Chart / ChartSeries / ChartCategoryAxis / ChartValueAxis β Powers the live viewership area chart in the Control Room.
Input: Drives the main session search bar.
Button: Handles the personalise drawer trigger and clearing searches.
Chip: Controls the track filter bar (React / JS / DevOps / UX / Perf / AI).
ProgressBar: Visualizes room capacity limits on live session cards.
Notification: Notification / NotificationGroup β Deploys the 2-minute session start toasts.
Tooltip: Reveals deep session descriptions on hover within the EPG timeline.
This project is a fully functional prototype driven by simulated real-time data. It demonstrates what the platform looks like when integrated with real conference infrastructure (ticketing systems, live stream URLs, and session CMS).
Note on the AI Engine: The recommendation engine is entirely deterministic. It matches session tracks against attendee levels to produce a 0β99 score and label (Top Pick / Great Match / Good Fit / Explore). It requires no external API calls and works completely offline.
CMS Integration: Automatically pull live stream URLs and session data from platforms like Sessionize or Hopin.
Live Ticketing: Connect to real ticketing and check-in data for perfectly accurate physical attendance numbers.
Automated VOD: Auto-populate Video-on-Demand links the moment recordings are published for a seamless "missed it? watch it" loop.
Smart Push Notifications: Alert users instantly via push notifications for any sessions the AI marks as a Top Pick.
Organiser Actions: Allow managers to send direct room alerts or flag capacity issues straight to venue staff from the Control Room dashboard.
Name: Tejasvi Nagendrappa
Profile: github.com/tejasvi-nagendrappa