HotelDigiBot — AI Chat Assistant for Hotels (SaaS)
- Category
- SaaS Platform
- Year
- 2026
- Technologies used
- Next.js 16, React 19, TypeScript, Tailwind CSS 4, PostgreSQL 16, pgvector, Prisma, Redis, BullMQ, NextAuth, Docker Compose, nginx, OpenAI API, Meta Graph API
- Live address
- hoteldigibot.net

Screenshots
Project overview
HotelDigiBot is a multi-tenant SaaS product that lets hotels and hospitality businesses answer guest questions with AI. Clinics, restaurants and similar businesses can use the same platform. Each business manages its own bot, content and inbox as a separate tenant.
The bot answers from the business's own content rather than general knowledge. Sites are crawled from a sitemap or URL list, and PDF and DOCX files can be uploaded. Content is turned into embeddings stored with pgvector, and relevant passages are found through semantic search. OpenAI gpt-4o-mini handles chat and text-embedding-3-small produces the embeddings. Bots can answer in multiple languages; the panel UI is in Turkish and English.
Channels: an embeddable website chat widget with configurable colour, logo and position; WhatsApp and Instagram connected via Meta Embedded Signup; and Telegram. All conversations land in a shared inbox, and a conversation can be handed off to a human when needed. Answers can be rated with like/dislike feedback.
Leads are detected from conversations and tracked in a pipeline. The system also includes analytics with response times, email and Telegram notifications and digests, anti-abuse measures and rate limiting, super-admin tenant management and Google Tag Manager events. Subscription plans are Free, Growth, Pro and Agency, with payments through PayTR and Stripe.
The stack is Next.js 16, React 19, TypeScript and Tailwind 4, with PostgreSQL 16 + pgvector and Prisma in the data layer, Redis and BullMQ workers for background jobs, and NextAuth for authentication. It runs on Docker Compose behind nginx and integrates with the OpenAI API and Meta Graph API.
Key features
- Multi-tenant architecture with super-admin tenant management
- Embeddable website chat widget with configurable colour, logo and position
- WhatsApp and Instagram (via Meta Embedded Signup) plus Telegram channels
- Answers from the business's own content via sitemap/URL crawling and PDF, DOCX uploads
- Semantic search on pgvector embeddings; gpt-4o-mini and text-embedding-3-small
- Multi-language bots; panel UI in Turkish and English
- Shared inbox, human handoff and like/dislike feedback
- Lead detection from conversations with a pipeline
- Analytics including response times, email and Telegram notifications and digests
- Free, Growth, Pro and Agency plans; PayTR and Stripe payments
Technologies used
- Next.js 16
- React 19
- TypeScript
- Tailwind CSS 4
- PostgreSQL 16
- pgvector
- Prisma
- Redis
- BullMQ
- NextAuth
- Docker Compose
- nginx
- OpenAI API
- Meta Graph API
