ExAi | Devpost

Inspiration

Trade shows generate thousands of conversations, but most of that knowledge disappears once the event ends. Organizers struggle to measure engagement, exhibitors lose valuable leads, and attendees often miss the people and booths most relevant to them. We wanted to build an AI-native platform that doesn’t just manage events—it understands them. Our vision is to create an AI operating system for the events industry where every stakeholder has an intelligent assistant before, during, and after the event.

What it does

ExAi is an AI-powered trade show platform built for organizers, exhibitors, and attendees.

For organizers, it provides live event analytics, AI-generated executive insights, reports, visitor intelligence, and engagement tracking.

For exhibitors, it offers AI-powered booth intelligence, visitor analytics, lead management, knowledge bases, QR-enabled engagement, and AI-assisted conversations with attendees.

For attendees, it delivers a personalized event experience with booth discovery, AI assistance, and networking support.

The platform uses live analytics and AI to transform event data into actionable insights instead of static dashboards.

How we built it

We built ExAi as a full-stack AI-native application using a modern monorepo architecture.

Next.js for the web application
NestJS for backend APIs
Supabase for authentication, PostgreSQL, realtime services, and storage
Drizzle ORM for database management
NVIDIA-powered LLM integration for AI insights and report generation
Redis and BullMQ for background processing
TypeScript across the entire stack
QR-powered attendee interactions
AI knowledge retrieval using exhibitor documents and structured company information

We also built a complete demo simulation engine that generates realistic visitor activity, analytics, engagement metrics, and AI-powered recommendations.

Challenges we ran into

Building an AI-first platform required much more than connecting an LLM.

Some of our biggest challenges included:

Designing a scalable multi-tenant architecture
Building a realistic event simulation instead of relying on static demo data
Connecting analytics, AI, and simulation into a single data pipeline
Managing a complex monorepo deployment with shared packages
Creating AI insights that are grounded in real event data rather than generic responses
Balancing rapid feature development with production-level stability under hackathon time constraints

Accomplishments that we're proud of
Built an end-to-end AI-powered event platform rather than a simple prototype
Created separate intelligent experiences for organizers, exhibitors, and attendees
Developed a live analytics engine with dynamic event insights
Implemented AI-generated reports and executive recommendations
Built an interactive event simulation to power a realistic demo
Designed a scalable architecture that can evolve into a production platform
Integrated AI throughout the product instead of limiting it to a chatbot

What we learned

This project taught us that building an AI product is fundamentally different from adding AI to an existing application.

The hardest challenge wasn’t integrating an LLM—it was designing reliable data pipelines, grounding AI responses in real event data, and creating workflows where AI delivers meaningful business value. We also gained valuable experience in architecting scalable monorepos, real-time analytics, and AI-driven user experiences under tight deadlines.

What's next for ExAi

Our vision is to evolve ExAi into the AI Operating System for the global events industry.

Next, we plan to introduce:

AI Event Creator that generates complete events from a simple prompt
AI Booth Builder that creates exhibitor experiences from websites and brochures
AI Brand Studio for automated booth branding and marketing assets
AI Concierge that recommends booths, sessions, meetings, and networking opportunities for attendees
Autonomous AI agents for organizers that monitor events in real time and recommend operational decisions
Predictive analytics, lead scoring, multilingual AI assistants, and deeper CRM integrations

Our long-term goal is simple: every trade show should have an AI operating system that helps organizers run smarter events, exhibitors close more business, and attendees build more meaningful connections.