Customizable chatbot that learns from your content to deliver natural language customer support
ChatFabrica intelligently learns by scanning the web pages and PDFs you select, enabling seamless integration into your website. With ChatFabrica, you can instantly deploy a customer support chatbot tailored to your content, enhancing your user experience from day one.
It can be quickly deployed through a straightforward integration process, offering customizable responses that align with your brand. Its user-friendly interface fosters natural conversations, while the system continuously learns and evolves to enhance customer interactions over time.
If you'd like to go with a SaaS service instead, check ChatFabrica's SaaS offerings.
- Customer support: Give instant responses to customer queries, ensuring that your customers receive the assistance they need without delay.
- Sales: Build your dedicated sales assistant, enhancing customer engagement and boosting conversions. ChatFabrica can also provide personalized recommendations and offer tailored product and service recommendations.
- Marketing: Automate interactions to generate leads, build relationships with potential customers, and increase your conversion rates.
- Open source under AGPLv3 license
- Learn from your PDFs (reports, documents, faqs)
- Learn from your webpage (in full or in part)
- Customizable responses
- Custom tone (e.g fun, serious, or academic)
- Chat log records and conversation histories
- Multi language support
Roadmap:
- Real support agent integration (live chat)
- API usage
- Gather and assess leads
- Zapier integration
- New AI models
- Facebook and Instagram integration
- Frontend: TypeScript, NextJS and Tailwind CSS.
- Backend: TypeScript, Nestjs, PrismaORM, MongoDB and OpenAI
This project consists of three main parts: client (Next.js), server (Nest.js) and crawler (Cheerio-Crawlee). Installation steps for each part are given below.
- Node.js (v14 or later recommended)
- npm
- Git
git clone https://github.com/bluewave-labs/bluewave-chatfabrica.git
cd bluewave-chatfabricacd Client
npm install
Create the .env file and fill in the necessary variables:
NEXT_PUBLIC_SENTRY_DSN="https://XXX.ingest.us.sentry.io/XXX"# DSN for Sentry bug tracking
AUTH_SECRET="anysecretkey" # Secret key for authentication
NEXT_PUBLIC_URL=http://localhost:3000 # Frontend application URL
NEXT_PUBLIC_BACKEND_URL=http://localhost:8000/api/v1 # URL of the Backend API
LEMONSQUEEZY_API_KEY="From Lemon Squeezy Panel"
LEMONSQUEEZY_STORE_ID="From Lemon Squeezy Panel"
LEMON_SQUEEZY_WEBHOOK_SIGNATURE="From Lemon Squeezy Panel"
NEXT_PUBLIC_GOOGLE_ANALYTICS_ID="G-XXXXXXXXXX"
SENTRY_AUTH_TOKEN="From Sentry Panel"
cd Server
npm install
Create .env file and fill in necessary variables:
NODE_ENV=development
APP_PORT=8000
APP_NAME="ChatFabrica API"
API_PREFIX=api/v1
FRONTEND_DOMAIN=http://localhost:3000
BACKEND_DOMAIN=http://localhost:8000
CRAWL_API_URL="http://localhost:8080/load-url"
LEMONSQUEEZY_WEBHOOK_SECRET="From Lemon Squeezy Panel"
DATABASE_URL="mongodb+srv://<USERNAME>:<PASSWORD>@<DB-URL>/<DB-NAME>?retryWrites=true&w=majority"
OPENAI_API_KEY="sk-XXX"
JWT_SECRET="anysecretkey"
MAIL_HOST=smtp.zoho.com
MAIL_PORT=465
[email protected]
MAIL_PASSWORD=mailpassword
MAIL_IGNORE_TLS=true
MAIL_SECURE=false
MAIL_REQUIRE_TLS=false
[email protected]
MAIL_DEFAULT_NAME=Api
ENCRYPTION_KEY=anyencryptionkey
cd Crawler
npm install
Open separate terminal windows for each section and run the following commands:
cd Client
npm run dev
cd Server
npm run db:migrate
npm run start:dev
cd Crawler
npm run dev
The client will run at http://localhost:3000.
The server will run at http://localhost:8000.
The crawler will run at http://localhost:8080.
Make sure to fill in all .env files. These files contain important configuration information that is required for the application to run properly.
If you encounter any problems or need more information, please contact the project manager.