WeaveFlow AI - Multi-Website Grounded Chatbot & Lead Capture SaaS

A complete, production-ready SaaS platform built as a 1:1 equivalent of AnswerWeave, powered by Grounded RAG, Zero Hallucination Guardrails, AI Call-Prep Briefs, and Multi-Website Tenancy.

---

🌟 Key SaaS Capabilities

1. Multi-Website / Multi-Assistant Management:

   • Manage multiple client websites or separate brands from a single SaaS dashboard.
   • Each website assistant has an isolated knowledge base, custom branding, domain whitelist, and dedicated lead pipeline.

2. 1-Line Embeddable Chat Widget (widget.js):

   • Drop into any website (WordPress, Shopify, Webflow, React, HTML) using a single script tag:

     <script src="http://localhost:8000/static/widget.js" data-assistant-id="asst_default"></script>
     

   • Includes Voice Question Support (Microphone 🎤) via browser speech recognition.

3. Strict Zero-Hallucination Grounding:

   • Adheres exclusively to your verified website pages, PDFs, and FAQs.
   • Refuses to invent facts when answers are not documented, stating the limitation and prompting for contact info.
   • Displays verifiable inline source citations ([1], [2]).

4. Lead Capture & Signature AI Call-Prep Briefs:

   • Captures visitor leads directly in the chat stream.
   • Automatically generates an Executive AI Call-Prep Brief for every lead (intent level, pain points, and recommended talking points for the sales rep).
   • Export leads to CSV with 1 click or send real-time webhooks.

5. Multi-Engine AI Support:

   • SiliconFlow (Free Chinese DeepSeek-V3 & Qwen 2.5): Free credits on registration, high-speed OpenAI-compatible API.
   • DeepSeek Official: DeepSeek-V3 and DeepSeek-R1.
   • Alibaba Qwen: DashScope Qwen-plus / Qwen-turbo.
   • Groq (Free Llama 3.3): Instant setup with zero credit card required.
   • Google Gemini (Free tier): Gemini 2.5 / 1.5 Flash.
   • OpenAI: GPT-4o-mini.
   • Zero-API-Key Local Engine (Active out-of-the-box): 100% free, runs offline without any API key.

---

⚡ Deploying to Vercel via GitHub

This repository is pre-configured for instant serverless deployment on Vercel:

• api/index.py: Serverless ASGI entry point.
• vercel.json: Automated URL rewrites for all REST routes and static assets.
• requirements.txt: Python package specification.

Step 1: Push Code to GitHub

git init
git add .
git commit -m "Initial commit of WeaveFlow AI"
git branch -M main
git remote add origin https://github.com/YOUR_USERNAME/YOUR_REPOSITORY.git
git push -u origin main

Step 2: Import into Vercel

1. Go to vercel.com/new and log in with your GitHub account.
2. Select your newly pushed repository and click Import.
3. Leave Framework Preset as Other (Vercel automatically detects Python serverless functions via api/index.py and vercel.json).
4. (Optional) In Environment Variables, add:
   • SILICONFLOW_API_KEY: (Optional free Chinese model key)
   • GEMINI_API_KEY: (Optional Google Gemini key)
   • GROQ_API_KEY: (Optional Groq key)
5. Click Deploy!

Within ~60 seconds, your WeaveFlow AI SaaS platform will be live at https://your-project.vercel.app with free SSL!

---

🚀 Local Development

Run the server locally using Python 3:

py -3 run.py

Accessing the SaaS Applications

• SaaS Admin Console & Multi-Site Manager: http://127.0.0.1:8000
• Live Multi-Website Simulator: http://127.0.0.1:8000/demo
• Interactive OpenAPI Documentation: http://127.0.0.1:8000/docs